rocksdb/db/corruption_test.cc

1510 lines
52 KiB
C++
Raw Normal View History

// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
//
// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
#ifndef ROCKSDB_LITE
#include <fcntl.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <cinttypes>
#include "db/db_impl/db_impl.h"
#include "db/db_test_util.h"
#include "db/log_format.h"
#include "db/version_set.h"
#include "file/filename.h"
#include "port/stack_trace.h"
#include "rocksdb/cache.h"
#include "rocksdb/convenience.h"
#include "rocksdb/db.h"
#include "rocksdb/env.h"
#include "rocksdb/table.h"
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
#include "rocksdb/utilities/transaction_db.h"
#include "rocksdb/write_batch.h"
#include "table/block_based/block_based_table_builder.h"
#include "table/meta_blocks.h"
#include "table/mock_table.h"
#include "test_util/testharness.h"
#include "test_util/testutil.h"
#include "util/cast_util.h"
#include "util/random.h"
#include "util/string_util.h"
namespace ROCKSDB_NAMESPACE {
static constexpr int kValueSize = 1000;
namespace {
// A wrapper that allows injection of errors.
class ErrorEnv : public EnvWrapper {
public:
bool writable_file_error_;
int num_writable_file_errors_;
explicit ErrorEnv(Env* _target)
: EnvWrapper(_target),
writable_file_error_(false),
num_writable_file_errors_(0) {}
const char* Name() const override { return "ErrorEnv"; }
virtual Status NewWritableFile(const std::string& fname,
std::unique_ptr<WritableFile>* result,
const EnvOptions& soptions) override {
result->reset();
if (writable_file_error_) {
++num_writable_file_errors_;
return Status::IOError(fname, "fake error");
}
return target()->NewWritableFile(fname, result, soptions);
}
};
} // namespace
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
class CorruptionTest : public testing::Test {
public:
std::shared_ptr<Env> env_guard_;
ErrorEnv* env_;
std::string dbname_;
std::shared_ptr<Cache> tiny_cache_;
Options options_;
DB* db_;
CorruptionTest() {
// If LRU cache shard bit is smaller than 2 (or -1 which will automatically
// set it to 0), test SequenceNumberRecovery will fail, likely because of a
// bug in recovery code. Keep it 4 for now to make the test passes.
tiny_cache_ = NewLRUCache(100, 4);
Env* base_env = Env::Default();
EXPECT_OK(
test::CreateEnvFromSystem(ConfigOptions(), &base_env, &env_guard_));
EXPECT_NE(base_env, nullptr);
env_ = new ErrorEnv(base_env);
options_.wal_recovery_mode = WALRecoveryMode::kTolerateCorruptedTailRecords;
options_.env = env_;
dbname_ = test::PerThreadDBPath(env_, "corruption_test");
Status s = DestroyDB(dbname_, options_);
EXPECT_OK(s);
db_ = nullptr;
options_.create_if_missing = true;
BlockBasedTableOptions table_options;
table_options.block_size_deviation = 0; // make unit test pass for now
options_.table_factory.reset(NewBlockBasedTableFactory(table_options));
Reopen();
options_.create_if_missing = false;
}
~CorruptionTest() override {
SyncPoint::GetInstance()->DisableProcessing();
SyncPoint::GetInstance()->LoadDependency({});
SyncPoint::GetInstance()->ClearAllCallBacks();
delete db_;
db_ = nullptr;
if (getenv("KEEP_DB")) {
fprintf(stdout, "db is still at %s\n", dbname_.c_str());
} else {
Options opts;
opts.env = env_->target();
EXPECT_OK(DestroyDB(dbname_, opts));
}
delete env_;
}
void CloseDb() {
delete db_;
db_ = nullptr;
}
Status TryReopen(Options* options = nullptr) {
delete db_;
db_ = nullptr;
Options opt = (options ? *options : options_);
if (opt.env == Options().env) {
// If env is not overridden, replace it with ErrorEnv.
// Otherwise, the test already uses a non-default Env.
opt.env = env_;
}
opt.arena_block_size = 4096;
BlockBasedTableOptions table_options;
table_options.block_cache = tiny_cache_;
table_options.block_size_deviation = 0;
opt.table_factory.reset(NewBlockBasedTableFactory(table_options));
return DB::Open(opt, dbname_, &db_);
}
void Reopen(Options* options = nullptr) {
ASSERT_OK(TryReopen(options));
}
void RepairDB() {
delete db_;
db_ = nullptr;
ASSERT_OK(::ROCKSDB_NAMESPACE::RepairDB(dbname_, options_));
}
void Build(int n, int start, int flush_every) {
std::string key_space, value_space;
WriteBatch batch;
for (int i = 0; i < n; i++) {
if (flush_every != 0 && i != 0 && i % flush_every == 0) {
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
}
//if ((i % 100) == 0) fprintf(stderr, "@ %d of %d\n", i, n);
Slice key = Key(i + start, &key_space);
batch.Clear();
ASSERT_OK(batch.Put(key, Value(i + start, &value_space)));
ASSERT_OK(db_->Write(WriteOptions(), &batch));
}
}
void Build(int n, int flush_every = 0) { Build(n, 0, flush_every); }
void Check(int min_expected, int max_expected) {
uint64_t next_expected = 0;
uint64_t missed = 0;
int bad_keys = 0;
int bad_values = 0;
int correct = 0;
std::string value_space;
// Do not verify checksums. If we verify checksums then the
// db itself will raise errors because data is corrupted.
// Instead, we want the reads to be successful and this test
// will detect whether the appropriate corruptions have
2015-04-25 11:14:27 +02:00
// occurred.
Iterator* iter = db_->NewIterator(ReadOptions(false, true));
for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
ASSERT_OK(iter->status());
uint64_t key;
Slice in(iter->key());
if (!ConsumeDecimalNumber(&in, &key) ||
!in.empty() ||
key < next_expected) {
bad_keys++;
continue;
}
missed += (key - next_expected);
next_expected = key + 1;
if (iter->value() != Value(static_cast<int>(key), &value_space)) {
bad_values++;
} else {
correct++;
}
}
iter->status().PermitUncheckedError();
delete iter;
fprintf(stderr,
"expected=%d..%d; got=%d; bad_keys=%d; bad_values=%d; missed=%llu\n",
min_expected, max_expected, correct, bad_keys, bad_values,
static_cast<unsigned long long>(missed));
ASSERT_LE(min_expected, correct);
ASSERT_GE(max_expected, correct);
}
void Corrupt(FileType filetype, int offset, int bytes_to_corrupt) {
// Pick file to corrupt
std::vector<std::string> filenames;
ASSERT_OK(env_->GetChildren(dbname_, &filenames));
uint64_t number;
FileType type;
std::string fname;
int picked_number = -1;
for (size_t i = 0; i < filenames.size(); i++) {
if (ParseFileName(filenames[i], &number, &type) &&
type == filetype &&
static_cast<int>(number) > picked_number) { // Pick latest file
fname = dbname_ + "/" + filenames[i];
picked_number = static_cast<int>(number);
}
}
ASSERT_TRUE(!fname.empty()) << filetype;
ASSERT_OK(test::CorruptFile(env_, fname, offset, bytes_to_corrupt));
}
// corrupts exactly one file at level `level`. if no file found at level,
// asserts
void CorruptTableFileAtLevel(int level, int offset, int bytes_to_corrupt) {
std::vector<LiveFileMetaData> metadata;
db_->GetLiveFilesMetaData(&metadata);
for (const auto& m : metadata) {
if (m.level == level) {
ASSERT_OK(test::CorruptFile(env_, dbname_ + "/" + m.name, offset,
Fix many tests to run with MEM_ENV and ENCRYPTED_ENV; Introduce a MemoryFileSystem class (#7566) Summary: This PR does a few things: 1. The MockFileSystem class was split out from the MockEnv. This change would theoretically allow a MockFileSystem to be used by other Environments as well (if we created a means of constructing one). The MockFileSystem implements a FileSystem in its entirety and does not rely on any Wrapper implementation. 2. Make the RocksDB test suite work when MOCK_ENV=1 and ENCRYPTED_ENV=1 are set. To accomplish this, a few things were needed: - The tests that tried to use the "wrong" environment (Env::Default() instead of env_) were updated - The MockFileSystem was changed to support the features it was missing or mishandled (such as recursively deleting files in a directory or supporting renaming of a directory). 3. Updated the test framework to have a ROCKSDB_GTEST_SKIP macro. This can be used to flag tests that are skipped. Currently, this defaults to doing nothing (marks the test as SUCCESS) but will mark the tests as SKIPPED when RocksDB is upgraded to a version of gtest that supports this (gtest-1.10). I have run a full "make check" with MEM_ENV, ENCRYPTED_ENV, both, and neither under both MacOS and RedHat. A few tests were disabled/skipped for the MEM/ENCRYPTED cases. The error_handler_fs_test fails/hangs for MEM_ENV (presumably a timing problem) and I will introduce another PR/issue to track that problem. (I will also push a change to disable those tests soon). There is one more test in DBTest2 that also fails which I need to investigate or skip before this PR is merged. Theoretically, this PR should also allow the test suite to run against an Env loaded from the registry, though I do not have one to try it with currently. Finally, once this is accepted, it would be nice if there was a CircleCI job to run these tests on a checkin so this effort does not become stale. I do not know how to do that, so if someone could write that job, it would be appreciated :) Pull Request resolved: https://github.com/facebook/rocksdb/pull/7566 Reviewed By: zhichao-cao Differential Revision: D24408980 Pulled By: jay-zhuang fbshipit-source-id: 911b1554a4d0da06fd51feca0c090a4abdcb4a5f
2020-10-27 18:31:34 +01:00
bytes_to_corrupt));
return;
}
}
FAIL() << "no file found at level";
}
int Property(const std::string& name) {
std::string property;
int result;
if (db_->GetProperty(name, &property) &&
sscanf(property.c_str(), "%d", &result) == 1) {
return result;
} else {
return -1;
}
}
// Return the ith key
Slice Key(int i, std::string* storage) {
char buf[100];
snprintf(buf, sizeof(buf), "%016d", i);
storage->assign(buf, strlen(buf));
return Slice(*storage);
}
// Return the value to associate with the specified key
Slice Value(int k, std::string* storage) {
if (k == 0) {
// Ugh. Random seed of 0 used to produce no entropy. This code
// preserves the implementation that was in place when all of the
// magic values in this file were picked.
*storage = std::string(kValueSize, ' ');
} else {
Random r(k);
*storage = r.RandomString(kValueSize);
}
return Slice(*storage);
}
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
void GetSortedWalFiles(std::vector<uint64_t>& file_nums) {
std::vector<std::string> tmp_files;
ASSERT_OK(env_->GetChildren(dbname_, &tmp_files));
FileType type = kWalFile;
for (const auto& file : tmp_files) {
uint64_t number = 0;
if (ParseFileName(file, &number, &type) && type == kWalFile) {
file_nums.push_back(number);
}
}
std::sort(file_nums.begin(), file_nums.end());
}
void CorruptFileWithTruncation(FileType file, uint64_t number,
uint64_t bytes_to_truncate = 0) {
std::string path;
switch (file) {
case FileType::kWalFile:
path = LogFileName(dbname_, number);
break;
// TODO: Add other file types as this method is being used for those file
// types.
default:
return;
}
uint64_t old_size = 0;
ASSERT_OK(env_->GetFileSize(path, &old_size));
assert(old_size > bytes_to_truncate);
uint64_t new_size = old_size - bytes_to_truncate;
// If bytes_to_truncate == 0, it will do full truncation.
if (bytes_to_truncate == 0) {
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
new_size = 0;
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
}
ASSERT_OK(test::TruncateFile(env_, path, new_size));
}
};
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, Recovery) {
Build(100);
Check(100, 100);
#ifdef OS_WIN
// On Wndows OS Disk cache does not behave properly
// We do not call FlushBuffers on every Flush. If we do not close
// the log file prior to the corruption we end up with the first
// block not corrupted but only the second. However, under the debugger
// things work just fine but never pass when running normally
// For that reason people may want to run with unbuffered I/O. That option
// is not available for WAL though.
CloseDb();
#endif
Corrupt(kWalFile, 19, 1); // WriteBatch tag for first record
Corrupt(kWalFile, log::kBlockSize + 1000, 1); // Somewhere in second block
ASSERT_TRUE(!TryReopen().ok());
options_.paranoid_checks = false;
Reopen(&options_);
// The 64 records in the first two log blocks are completely lost.
Check(36, 36);
}
TEST_F(CorruptionTest, PostPITRCorruptionWALsRetained) {
// Repro for bug where WALs following the point-in-time recovery were not
// retained leading to the next recovery failing.
CloseDb();
options_.wal_recovery_mode = WALRecoveryMode::kPointInTimeRecovery;
const std::string test_cf_name = "test_cf";
std::vector<ColumnFamilyDescriptor> cf_descs;
cf_descs.emplace_back(kDefaultColumnFamilyName, ColumnFamilyOptions());
cf_descs.emplace_back(test_cf_name, ColumnFamilyOptions());
uint64_t log_num;
{
options_.create_missing_column_families = true;
std::vector<ColumnFamilyHandle*> cfhs;
ASSERT_OK(DB::Open(options_, dbname_, cf_descs, &cfhs, &db_));
assert(db_ != nullptr); // suppress false clang-analyze report
ASSERT_OK(db_->Put(WriteOptions(), cfhs[0], "k", "v"));
ASSERT_OK(db_->Put(WriteOptions(), cfhs[1], "k", "v"));
ASSERT_OK(db_->Put(WriteOptions(), cfhs[0], "k2", "v2"));
std::vector<uint64_t> file_nums;
GetSortedWalFiles(file_nums);
log_num = file_nums.back();
for (auto* cfh : cfhs) {
delete cfh;
}
CloseDb();
}
CorruptFileWithTruncation(FileType::kWalFile, log_num,
/*bytes_to_truncate=*/1);
{
// Recover "k" -> "v" for both CFs. "k2" -> "v2" is lost due to truncation.
options_.avoid_flush_during_recovery = true;
std::vector<ColumnFamilyHandle*> cfhs;
ASSERT_OK(DB::Open(options_, dbname_, cf_descs, &cfhs, &db_));
assert(db_ != nullptr); // suppress false clang-analyze report
// Flush one but not both CFs and write some data so there's a seqno gap
// between the PITR corruption and the next DB session's first WAL.
ASSERT_OK(db_->Put(WriteOptions(), cfhs[1], "k2", "v2"));
ASSERT_OK(db_->Flush(FlushOptions(), cfhs[1]));
for (auto* cfh : cfhs) {
delete cfh;
}
CloseDb();
}
// With the bug, this DB open would remove the WALs following the PITR
// corruption. Then, the next recovery would fail.
for (int i = 0; i < 2; ++i) {
std::vector<ColumnFamilyHandle*> cfhs;
ASSERT_OK(DB::Open(options_, dbname_, cf_descs, &cfhs, &db_));
assert(db_ != nullptr); // suppress false clang-analyze report
for (auto* cfh : cfhs) {
delete cfh;
}
CloseDb();
}
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, RecoverWriteError) {
env_->writable_file_error_ = true;
Status s = TryReopen();
ASSERT_TRUE(!s.ok());
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, NewFileErrorDuringWrite) {
// Do enough writing to force minor compaction
env_->writable_file_error_ = true;
const int num =
static_cast<int>(3 + (Options().write_buffer_size / kValueSize));
std::string value_storage;
Status s;
bool failed = false;
for (int i = 0; i < num; i++) {
WriteBatch batch;
ASSERT_OK(batch.Put("a", Value(100, &value_storage)));
s = db_->Write(WriteOptions(), &batch);
if (!s.ok()) {
failed = true;
}
ASSERT_TRUE(!failed || !s.ok());
}
ASSERT_TRUE(!s.ok());
ASSERT_GE(env_->num_writable_file_errors_, 1);
env_->writable_file_error_ = false;
Reopen();
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, TableFile) {
Build(100);
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr));
ASSERT_OK(dbi->TEST_CompactRange(1, nullptr, nullptr));
Corrupt(kTableFile, 100, 1);
Check(99, 99);
ASSERT_NOK(dbi->VerifyChecksum());
}
TEST_F(CorruptionTest, VerifyChecksumReadahead) {
Options options;
SpecialEnv senv(env_->target());
options.env = &senv;
// Disable block cache as we are going to check checksum for
// the same file twice and measure number of reads.
BlockBasedTableOptions table_options_no_bc;
table_options_no_bc.no_block_cache = true;
options.table_factory.reset(NewBlockBasedTableFactory(table_options_no_bc));
Reopen(&options);
Build(10000);
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr));
ASSERT_OK(dbi->TEST_CompactRange(1, nullptr, nullptr));
senv.count_random_reads_ = true;
senv.random_read_counter_.Reset();
ASSERT_OK(dbi->VerifyChecksum());
// Make sure the counter is enabled.
ASSERT_GT(senv.random_read_counter_.Read(), 0);
// The SST file is about 10MB. Default readahead size is 256KB.
// Give a conservative 20 reads for metadata blocks, The number
// of random reads should be within 10 MB / 256KB + 20 = 60.
ASSERT_LT(senv.random_read_counter_.Read(), 60);
senv.random_read_bytes_counter_ = 0;
ReadOptions ro;
ro.readahead_size = size_t{32 * 1024};
ASSERT_OK(dbi->VerifyChecksum(ro));
// The SST file is about 10MB. We set readahead size to 32KB.
// Give 0 to 20 reads for metadata blocks, and allow real read
// to range from 24KB to 48KB. The lower bound would be:
// 10MB / 48KB + 0 = 213
// The higher bound is
// 10MB / 24KB + 20 = 447.
ASSERT_GE(senv.random_read_counter_.Read(), 213);
ASSERT_LE(senv.random_read_counter_.Read(), 447);
// Test readahead shouldn't break mmap mode (where it should be
// disabled).
options.allow_mmap_reads = true;
Reopen(&options);
dbi = static_cast<DBImpl*>(db_);
ASSERT_OK(dbi->VerifyChecksum(ro));
CloseDb();
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, TableFileIndexData) {
Options options;
// very big, we'll trigger flushes manually
options.write_buffer_size = 100 * 1024 * 1024;
Reopen(&options);
// build 2 tables, flush at 5000
Build(10000, 5000);
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
// corrupt an index block of an entire file
Corrupt(kTableFile, -2000, 500);
options.paranoid_checks = false;
Reopen(&options);
dbi = static_cast_with_check<DBImpl>(db_);
Change and clarify the relationship between Valid(), status() and Seek*() for all iterators. Also fix some bugs Summary: Before this PR, Iterator/InternalIterator may simultaneously have non-ok status() and Valid() = true. That state means that the last operation failed, but the iterator is nevertheless positioned on some unspecified record. Likely intended uses of that are: * If some sst files are corrupted, a normal iterator can be used to read the data from files that are not corrupted. * When using read_tier = kBlockCacheTier, read the data that's in block cache, skipping over the data that is not. However, this behavior wasn't documented well (and until recently the wiki on github had misleading incorrect information). In the code there's a lot of confusion about the relationship between status() and Valid(), and about whether Seek()/SeekToLast()/etc reset the status or not. There were a number of bugs caused by this confusion, both inside rocksdb and in the code that uses rocksdb (including ours). This PR changes the convention to: * If status() is not ok, Valid() always returns false. * Any seek operation resets status. (Before the PR, it depended on iterator type and on particular error.) This does sacrifice the two use cases listed above, but siying said it's ok. Overview of the changes: * A commit that adds missing status checks in MergingIterator. This fixes a bug that actually affects us, and we need it fixed. `DBIteratorTest.NonBlockingIterationBugRepro` explains the scenario. * Changes to lots of iterator types to make all of them conform to the new convention. Some bug fixes along the way. By far the biggest changes are in DBIter, which is a big messy piece of code; I tried to make it less big and messy but mostly failed. * A stress-test for DBIter, to gain some confidence that I didn't break it. It does a few million random operations on the iterator, while occasionally modifying the underlying data (like ForwardIterator does) and occasionally returning non-ok status from internal iterator. To find the iterator types that needed changes I searched for "public .*Iterator" in the code. Here's an overview of all 27 iterator types: Iterators that didn't need changes: * status() is always ok(), or Valid() is always false: MemTableIterator, ModelIter, TestIterator, KVIter (2 classes with this name anonymous namespaces), LoggingForwardVectorIterator, VectorIterator, MockTableIterator, EmptyIterator, EmptyInternalIterator. * Thin wrappers that always pass through Valid() and status(): ArenaWrappedDBIter, TtlIterator, InternalIteratorFromIterator. Iterators with changes (see inline comments for details): * DBIter - an overhaul: - It used to silently skip corrupted keys (`FindParseableKey()`), which seems dangerous. This PR makes it just stop immediately after encountering a corrupted key, just like it would for other kinds of corruption. Let me know if there was actually some deeper meaning in this behavior and I should put it back. - It had a few code paths silently discarding subiterator's status. The stress test caught a few. - The backwards iteration code path was expecting the internal iterator's set of keys to be immutable. It's probably always true in practice at the moment, since ForwardIterator doesn't support backwards iteration, but this PR fixes it anyway. See added DBIteratorTest.ReverseToForwardBug for an example. - Some parts of backwards iteration code path even did things like `assert(iter_->Valid())` after a seek, which is never a safe assumption. - It used to not reset status on seek for some types of errors. - Some simplifications and better comments. - Some things got more complicated from the added error handling. I'm open to ideas for how to make it nicer. * MergingIterator - check status after every operation on every subiterator, and in some places assert that valid subiterators have ok status. * ForwardIterator - changed to the new convention, also slightly simplified. * ForwardLevelIterator - fixed some bugs and simplified. * LevelIterator - simplified. * TwoLevelIterator - changed to the new convention. Also fixed a bug that would make SeekForPrev() sometimes silently ignore errors from first_level_iter_. * BlockBasedTableIterator - minor changes. * BlockIter - replaced `SetStatus()` with `Invalidate()` to make sure non-ok BlockIter is always invalid. * PlainTableIterator - some seeks used to not reset status. * CuckooTableIterator - tiny code cleanup. * ManagedIterator - fixed some bugs. * BaseDeltaIterator - changed to the new convention and fixed a bug. * BlobDBIterator - seeks used to not reset status. * KeyConvertingIterator - some small change. Closes https://github.com/facebook/rocksdb/pull/3810 Differential Revision: D7888019 Pulled By: al13n321 fbshipit-source-id: 4aaf6d3421c545d16722a815b2fa2e7912bc851d
2018-05-17 11:44:14 +02:00
// one full file may be readable, since only one was corrupted
// the other file should be fully non-readable, since index was corrupted
Change and clarify the relationship between Valid(), status() and Seek*() for all iterators. Also fix some bugs Summary: Before this PR, Iterator/InternalIterator may simultaneously have non-ok status() and Valid() = true. That state means that the last operation failed, but the iterator is nevertheless positioned on some unspecified record. Likely intended uses of that are: * If some sst files are corrupted, a normal iterator can be used to read the data from files that are not corrupted. * When using read_tier = kBlockCacheTier, read the data that's in block cache, skipping over the data that is not. However, this behavior wasn't documented well (and until recently the wiki on github had misleading incorrect information). In the code there's a lot of confusion about the relationship between status() and Valid(), and about whether Seek()/SeekToLast()/etc reset the status or not. There were a number of bugs caused by this confusion, both inside rocksdb and in the code that uses rocksdb (including ours). This PR changes the convention to: * If status() is not ok, Valid() always returns false. * Any seek operation resets status. (Before the PR, it depended on iterator type and on particular error.) This does sacrifice the two use cases listed above, but siying said it's ok. Overview of the changes: * A commit that adds missing status checks in MergingIterator. This fixes a bug that actually affects us, and we need it fixed. `DBIteratorTest.NonBlockingIterationBugRepro` explains the scenario. * Changes to lots of iterator types to make all of them conform to the new convention. Some bug fixes along the way. By far the biggest changes are in DBIter, which is a big messy piece of code; I tried to make it less big and messy but mostly failed. * A stress-test for DBIter, to gain some confidence that I didn't break it. It does a few million random operations on the iterator, while occasionally modifying the underlying data (like ForwardIterator does) and occasionally returning non-ok status from internal iterator. To find the iterator types that needed changes I searched for "public .*Iterator" in the code. Here's an overview of all 27 iterator types: Iterators that didn't need changes: * status() is always ok(), or Valid() is always false: MemTableIterator, ModelIter, TestIterator, KVIter (2 classes with this name anonymous namespaces), LoggingForwardVectorIterator, VectorIterator, MockTableIterator, EmptyIterator, EmptyInternalIterator. * Thin wrappers that always pass through Valid() and status(): ArenaWrappedDBIter, TtlIterator, InternalIteratorFromIterator. Iterators with changes (see inline comments for details): * DBIter - an overhaul: - It used to silently skip corrupted keys (`FindParseableKey()`), which seems dangerous. This PR makes it just stop immediately after encountering a corrupted key, just like it would for other kinds of corruption. Let me know if there was actually some deeper meaning in this behavior and I should put it back. - It had a few code paths silently discarding subiterator's status. The stress test caught a few. - The backwards iteration code path was expecting the internal iterator's set of keys to be immutable. It's probably always true in practice at the moment, since ForwardIterator doesn't support backwards iteration, but this PR fixes it anyway. See added DBIteratorTest.ReverseToForwardBug for an example. - Some parts of backwards iteration code path even did things like `assert(iter_->Valid())` after a seek, which is never a safe assumption. - It used to not reset status on seek for some types of errors. - Some simplifications and better comments. - Some things got more complicated from the added error handling. I'm open to ideas for how to make it nicer. * MergingIterator - check status after every operation on every subiterator, and in some places assert that valid subiterators have ok status. * ForwardIterator - changed to the new convention, also slightly simplified. * ForwardLevelIterator - fixed some bugs and simplified. * LevelIterator - simplified. * TwoLevelIterator - changed to the new convention. Also fixed a bug that would make SeekForPrev() sometimes silently ignore errors from first_level_iter_. * BlockBasedTableIterator - minor changes. * BlockIter - replaced `SetStatus()` with `Invalidate()` to make sure non-ok BlockIter is always invalid. * PlainTableIterator - some seeks used to not reset status. * CuckooTableIterator - tiny code cleanup. * ManagedIterator - fixed some bugs. * BaseDeltaIterator - changed to the new convention and fixed a bug. * BlobDBIterator - seeks used to not reset status. * KeyConvertingIterator - some small change. Closes https://github.com/facebook/rocksdb/pull/3810 Differential Revision: D7888019 Pulled By: al13n321 fbshipit-source-id: 4aaf6d3421c545d16722a815b2fa2e7912bc851d
2018-05-17 11:44:14 +02:00
Check(0, 5000);
ASSERT_NOK(dbi->VerifyChecksum());
// In paranoid mode, the db cannot be opened due to the corrupted file.
ASSERT_TRUE(TryReopen().IsCorruption());
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, MissingDescriptor) {
Build(1000);
RepairDB();
Reopen();
Check(1000, 1000);
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, SequenceNumberRecovery) {
ASSERT_OK(db_->Put(WriteOptions(), "foo", "v1"));
ASSERT_OK(db_->Put(WriteOptions(), "foo", "v2"));
ASSERT_OK(db_->Put(WriteOptions(), "foo", "v3"));
ASSERT_OK(db_->Put(WriteOptions(), "foo", "v4"));
ASSERT_OK(db_->Put(WriteOptions(), "foo", "v5"));
RepairDB();
Reopen();
std::string v;
ASSERT_OK(db_->Get(ReadOptions(), "foo", &v));
ASSERT_EQ("v5", v);
// Write something. If sequence number was not recovered properly,
// it will be hidden by an earlier write.
ASSERT_OK(db_->Put(WriteOptions(), "foo", "v6"));
ASSERT_OK(db_->Get(ReadOptions(), "foo", &v));
ASSERT_EQ("v6", v);
Reopen();
ASSERT_OK(db_->Get(ReadOptions(), "foo", &v));
ASSERT_EQ("v6", v);
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, CorruptedDescriptor) {
ASSERT_OK(db_->Put(WriteOptions(), "foo", "hello"));
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr));
Corrupt(kDescriptorFile, 0, 1000);
Status s = TryReopen();
ASSERT_TRUE(!s.ok());
RepairDB();
Reopen();
std::string v;
ASSERT_OK(db_->Get(ReadOptions(), "foo", &v));
ASSERT_EQ("hello", v);
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, CompactionInputError) {
Options options;
options.env = env_;
Reopen(&options);
Build(10);
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr));
ASSERT_OK(dbi->TEST_CompactRange(1, nullptr, nullptr));
ASSERT_EQ(1, Property("rocksdb.num-files-at-level2"));
Corrupt(kTableFile, 100, 1);
Check(9, 9);
ASSERT_NOK(dbi->VerifyChecksum());
// Force compactions by writing lots of values
Build(10000);
Check(10000, 10000);
ASSERT_NOK(dbi->VerifyChecksum());
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, CompactionInputErrorParanoid) {
Options options;
options.env = env_;
options.paranoid_checks = true;
options.write_buffer_size = 131072;
options.max_write_buffer_number = 2;
Reopen(&options);
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
// Fill levels >= 1
for (int level = 1; level < dbi->NumberLevels(); level++) {
ASSERT_OK(dbi->Put(WriteOptions(), "", "begin"));
ASSERT_OK(dbi->Put(WriteOptions(), "~", "end"));
ASSERT_OK(dbi->TEST_FlushMemTable());
for (int comp_level = 0; comp_level < dbi->NumberLevels() - level;
++comp_level) {
ASSERT_OK(dbi->TEST_CompactRange(comp_level, nullptr, nullptr));
}
}
Reopen(&options);
dbi = static_cast_with_check<DBImpl>(db_);
Build(10);
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(dbi->TEST_WaitForCompact());
ASSERT_EQ(1, Property("rocksdb.num-files-at-level0"));
CorruptTableFileAtLevel(0, 100, 1);
Check(9, 9);
ASSERT_NOK(dbi->VerifyChecksum());
// Write must eventually fail because of corrupted table
Status s;
std::string tmp1, tmp2;
bool failed = false;
for (int i = 0; i < 10000; i++) {
s = db_->Put(WriteOptions(), Key(i, &tmp1), Value(i, &tmp2));
if (!s.ok()) {
failed = true;
}
// if one write failed, every subsequent write must fail, too
ASSERT_TRUE(!failed || !s.ok()) << "write did not fail in a corrupted db";
}
ASSERT_TRUE(!s.ok()) << "write did not fail in corrupted paranoid db";
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, UnrelatedKeys) {
Build(10);
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
Corrupt(kTableFile, 100, 1);
ASSERT_NOK(dbi->VerifyChecksum());
std::string tmp1, tmp2;
ASSERT_OK(db_->Put(WriteOptions(), Key(1000, &tmp1), Value(1000, &tmp2)));
std::string v;
ASSERT_OK(db_->Get(ReadOptions(), Key(1000, &tmp1), &v));
ASSERT_EQ(Value(1000, &tmp2).ToString(), v);
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(db_->Get(ReadOptions(), Key(1000, &tmp1), &v));
ASSERT_EQ(Value(1000, &tmp2).ToString(), v);
}
TEST_F(CorruptionTest, RangeDeletionCorrupted) {
ASSERT_OK(
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "b"));
ASSERT_OK(db_->Flush(FlushOptions()));
std::vector<LiveFileMetaData> metadata;
db_->GetLiveFilesMetaData(&metadata);
ASSERT_EQ(static_cast<size_t>(1), metadata.size());
std::string filename = dbname_ + metadata[0].name;
FileOptions file_opts;
const auto& fs = options_.env->GetFileSystem();
std::unique_ptr<RandomAccessFileReader> file_reader;
ASSERT_OK(RandomAccessFileReader::Create(fs, filename, file_opts,
&file_reader, nullptr));
uint64_t file_size;
ASSERT_OK(
fs->GetFileSize(filename, file_opts.io_options, &file_size, nullptr));
BlockHandle range_del_handle;
Improve / clean up meta block code & integrity (#9163) Summary: * Checksums are now checked on meta blocks unless specifically suppressed or not applicable (e.g. plain table). (Was other way around.) This means a number of cases that were not checking checksums now are, including direct read TableProperties in Version::GetTableProperties (fixed in meta_blocks ReadTableProperties), reading any block from PersistentCache (fixed in BlockFetcher), read TableProperties in SstFileDumper (ldb/sst_dump/BackupEngine) before table reader open, maybe more. * For that to work, I moved the global_seqno+TableProperties checksum logic to the shared table/ code, because that is used by many utilies such as SstFileDumper. * Also for that to work, we have to know when we're dealing with a block that has a checksum (trailer), so added that capability to Footer based on magic number, and from there BlockFetcher. * Knowledge of trailer presence has also fixed a problem where other table formats were reading blocks including bytes for a non-existant trailer--and awkwardly kind-of not using them, e.g. no shared code checking checksums. (BlockFetcher compression type was populated incorrectly.) Now we only read what is needed. * Minimized code duplication and differing/incompatible/awkward abstractions in meta_blocks.{cc,h} (e.g. SeekTo in metaindex block without parsing block handle) * Moved some meta block handling code from table_properties*.* * Moved some code specific to block-based table from shared table/ code to BlockBasedTable class. The checksum stuff means we can't completely separate it, but things that don't need to be in shared table/ code should not be. * Use unique_ptr rather than raw ptr in more places. (Note: you can std::move from unique_ptr to shared_ptr.) Without enhancements to GetPropertiesOfAllTablesTest (see below), net reduction of roughly 100 lines of code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/9163 Test Plan: existing tests and * Enhanced DBTablePropertiesTest.GetPropertiesOfAllTablesTest to verify that checksums are now checked on direct read of table properties by TableCache (new test would fail before this change) * Also enhanced DBTablePropertiesTest.GetPropertiesOfAllTablesTest to test putting table properties under old meta name * Also generally enhanced that same test to actually test what it was supposed to be testing already, by kicking things out of table cache when we don't want them there. Reviewed By: ajkr, mrambacher Differential Revision: D32514757 Pulled By: pdillinger fbshipit-source-id: 507964b9311d186ae8d1131182290cbd97a99fa9
2021-11-18 20:42:12 +01:00
ASSERT_OK(FindMetaBlockInFile(
file_reader.get(), file_size, kBlockBasedTableMagicNumber,
ImmutableOptions(options_), kRangeDelBlockName, &range_del_handle));
ASSERT_OK(TryReopen());
ASSERT_OK(test::CorruptFile(env_, filename,
Fix many tests to run with MEM_ENV and ENCRYPTED_ENV; Introduce a MemoryFileSystem class (#7566) Summary: This PR does a few things: 1. The MockFileSystem class was split out from the MockEnv. This change would theoretically allow a MockFileSystem to be used by other Environments as well (if we created a means of constructing one). The MockFileSystem implements a FileSystem in its entirety and does not rely on any Wrapper implementation. 2. Make the RocksDB test suite work when MOCK_ENV=1 and ENCRYPTED_ENV=1 are set. To accomplish this, a few things were needed: - The tests that tried to use the "wrong" environment (Env::Default() instead of env_) were updated - The MockFileSystem was changed to support the features it was missing or mishandled (such as recursively deleting files in a directory or supporting renaming of a directory). 3. Updated the test framework to have a ROCKSDB_GTEST_SKIP macro. This can be used to flag tests that are skipped. Currently, this defaults to doing nothing (marks the test as SUCCESS) but will mark the tests as SKIPPED when RocksDB is upgraded to a version of gtest that supports this (gtest-1.10). I have run a full "make check" with MEM_ENV, ENCRYPTED_ENV, both, and neither under both MacOS and RedHat. A few tests were disabled/skipped for the MEM/ENCRYPTED cases. The error_handler_fs_test fails/hangs for MEM_ENV (presumably a timing problem) and I will introduce another PR/issue to track that problem. (I will also push a change to disable those tests soon). There is one more test in DBTest2 that also fails which I need to investigate or skip before this PR is merged. Theoretically, this PR should also allow the test suite to run against an Env loaded from the registry, though I do not have one to try it with currently. Finally, once this is accepted, it would be nice if there was a CircleCI job to run these tests on a checkin so this effort does not become stale. I do not know how to do that, so if someone could write that job, it would be appreciated :) Pull Request resolved: https://github.com/facebook/rocksdb/pull/7566 Reviewed By: zhichao-cao Differential Revision: D24408980 Pulled By: jay-zhuang fbshipit-source-id: 911b1554a4d0da06fd51feca0c090a4abdcb4a5f
2020-10-27 18:31:34 +01:00
static_cast<int>(range_del_handle.offset()), 1));
ASSERT_TRUE(TryReopen().IsCorruption());
}
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
TEST_F(CorruptionTest, FileSystemStateCorrupted) {
for (int iter = 0; iter < 2; ++iter) {
Options options;
options.env = env_;
options.paranoid_checks = true;
options.create_if_missing = true;
Reopen(&options);
Build(10);
ASSERT_OK(db_->Flush(FlushOptions()));
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
std::vector<LiveFileMetaData> metadata;
dbi->GetLiveFilesMetaData(&metadata);
ASSERT_GT(metadata.size(), 0);
std::string filename = dbname_ + metadata[0].name;
delete db_;
2014-03-20 22:37:30 +01:00
db_ = nullptr;
if (iter == 0) { // corrupt file size
std::unique_ptr<WritableFile> file;
ASSERT_OK(env_->NewWritableFile(filename, &file, EnvOptions()));
ASSERT_OK(file->Append(Slice("corrupted sst")));
file.reset();
Status x = TryReopen(&options);
ASSERT_TRUE(x.IsCorruption());
} else { // delete the file
ASSERT_OK(env_->DeleteFile(filename));
Status x = TryReopen(&options);
ASSERT_TRUE(x.IsCorruption());
}
ASSERT_OK(DestroyDB(dbname_, options_));
}
}
static const auto& corruption_modes = {
mock::MockTableFactory::kCorruptNone, mock::MockTableFactory::kCorruptKey,
mock::MockTableFactory::kCorruptValue,
mock::MockTableFactory::kCorruptReorderKey};
TEST_F(CorruptionTest, ParanoidFileChecksOnFlush) {
Options options;
options.env = env_;
options.check_flush_compaction_key_order = false;
options.paranoid_file_checks = true;
options.create_if_missing = true;
Status s;
for (const auto& mode : corruption_modes) {
delete db_;
db_ = nullptr;
s = DestroyDB(dbname_, options);
ASSERT_OK(s);
std::shared_ptr<mock::MockTableFactory> mock =
std::make_shared<mock::MockTableFactory>();
options.table_factory = mock;
mock->SetCorruptionMode(mode);
ASSERT_OK(DB::Open(options, dbname_, &db_));
assert(db_ != nullptr); // suppress false clang-analyze report
Build(10);
s = db_->Flush(FlushOptions());
if (mode == mock::MockTableFactory::kCorruptNone) {
ASSERT_OK(s);
} else {
ASSERT_NOK(s);
}
}
}
TEST_F(CorruptionTest, ParanoidFileChecksOnCompact) {
Options options;
options.env = env_;
options.paranoid_file_checks = true;
options.create_if_missing = true;
options.check_flush_compaction_key_order = false;
Status s;
for (const auto& mode : corruption_modes) {
delete db_;
db_ = nullptr;
s = DestroyDB(dbname_, options);
std::shared_ptr<mock::MockTableFactory> mock =
std::make_shared<mock::MockTableFactory>();
options.table_factory = mock;
ASSERT_OK(DB::Open(options, dbname_, &db_));
assert(db_ != nullptr); // suppress false clang-analyze report
Build(100, 2);
// ASSERT_OK(db_->Flush(FlushOptions()));
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
mock->SetCorruptionMode(mode);
s = dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true);
if (mode == mock::MockTableFactory::kCorruptNone) {
ASSERT_OK(s);
} else {
ASSERT_NOK(s);
}
}
}
TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeFirst) {
Options options;
options.env = env_;
options.check_flush_compaction_key_order = false;
options.paranoid_file_checks = true;
options.create_if_missing = true;
for (bool do_flush : {true, false}) {
delete db_;
db_ = nullptr;
ASSERT_OK(DestroyDB(dbname_, options));
ASSERT_OK(DB::Open(options, dbname_, &db_));
std::string start, end;
assert(db_ != nullptr); // suppress false clang-analyze report
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(3, &start), Key(7, &end)));
auto snap = db_->GetSnapshot();
ASSERT_NE(snap, nullptr);
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(8, &start), Key(9, &end)));
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(2, &start), Key(5, &end)));
Build(10);
if (do_flush) {
ASSERT_OK(db_->Flush(FlushOptions()));
} else {
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true));
}
db_->ReleaseSnapshot(snap);
}
}
TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRange) {
Options options;
options.env = env_;
options.check_flush_compaction_key_order = false;
options.paranoid_file_checks = true;
options.create_if_missing = true;
for (bool do_flush : {true, false}) {
delete db_;
db_ = nullptr;
ASSERT_OK(DestroyDB(dbname_, options));
ASSERT_OK(DB::Open(options, dbname_, &db_));
assert(db_ != nullptr); // suppress false clang-analyze report
Build(10, 0, 0);
std::string start, end;
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(5, &start), Key(15, &end)));
auto snap = db_->GetSnapshot();
ASSERT_NE(snap, nullptr);
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(8, &start), Key(9, &end)));
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(12, &start), Key(17, &end)));
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(2, &start), Key(4, &end)));
Build(10, 10, 0);
if (do_flush) {
ASSERT_OK(db_->Flush(FlushOptions()));
} else {
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true));
}
db_->ReleaseSnapshot(snap);
}
}
TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeLast) {
Options options;
options.env = env_;
options.check_flush_compaction_key_order = false;
options.paranoid_file_checks = true;
options.create_if_missing = true;
for (bool do_flush : {true, false}) {
delete db_;
db_ = nullptr;
ASSERT_OK(DestroyDB(dbname_, options));
ASSERT_OK(DB::Open(options, dbname_, &db_));
assert(db_ != nullptr); // suppress false clang-analyze report
std::string start, end;
Build(10);
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(3, &start), Key(7, &end)));
auto snap = db_->GetSnapshot();
ASSERT_NE(snap, nullptr);
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(6, &start), Key(8, &end)));
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(2, &start), Key(5, &end)));
if (do_flush) {
ASSERT_OK(db_->Flush(FlushOptions()));
} else {
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true));
}
db_->ReleaseSnapshot(snap);
}
}
TEST_F(CorruptionTest, LogCorruptionErrorsInCompactionIterator) {
Options options;
options.env = env_;
options.create_if_missing = true;
options.allow_data_in_errors = true;
auto mode = mock::MockTableFactory::kCorruptKey;
delete db_;
db_ = nullptr;
ASSERT_OK(DestroyDB(dbname_, options));
std::shared_ptr<mock::MockTableFactory> mock =
std::make_shared<mock::MockTableFactory>();
mock->SetCorruptionMode(mode);
options.table_factory = mock;
ASSERT_OK(DB::Open(options, dbname_, &db_));
assert(db_ != nullptr); // suppress false clang-analyze report
Build(100, 2);
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
Status s = dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true);
ASSERT_NOK(s);
ASSERT_TRUE(s.IsCorruption());
}
TEST_F(CorruptionTest, CompactionKeyOrderCheck) {
Options options;
options.env = env_;
options.paranoid_file_checks = false;
options.create_if_missing = true;
options.check_flush_compaction_key_order = false;
delete db_;
db_ = nullptr;
ASSERT_OK(DestroyDB(dbname_, options));
std::shared_ptr<mock::MockTableFactory> mock =
std::make_shared<mock::MockTableFactory>();
options.table_factory = mock;
ASSERT_OK(DB::Open(options, dbname_, &db_));
assert(db_ != nullptr); // suppress false clang-analyze report
mock->SetCorruptionMode(mock::MockTableFactory::kCorruptReorderKey);
Build(100, 2);
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
ASSERT_OK(dbi->TEST_FlushMemTable());
mock->SetCorruptionMode(mock::MockTableFactory::kCorruptNone);
ASSERT_OK(db_->SetOptions({{"check_flush_compaction_key_order", "true"}}));
ASSERT_NOK(dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true));
}
TEST_F(CorruptionTest, FlushKeyOrderCheck) {
Options options;
options.env = env_;
options.paranoid_file_checks = false;
options.create_if_missing = true;
ASSERT_OK(db_->SetOptions({{"check_flush_compaction_key_order", "true"}}));
ASSERT_OK(db_->Put(WriteOptions(), "foo1", "v1"));
ASSERT_OK(db_->Put(WriteOptions(), "foo2", "v1"));
ASSERT_OK(db_->Put(WriteOptions(), "foo3", "v1"));
ASSERT_OK(db_->Put(WriteOptions(), "foo4", "v1"));
int cnt = 0;
// Generate some out of order keys from the memtable
SyncPoint::GetInstance()->SetCallBack(
"MemTableIterator::Next:0", [&](void* arg) {
MemTableRep::Iterator* mem_iter =
static_cast<MemTableRep::Iterator*>(arg);
if (++cnt == 3) {
mem_iter->Prev();
mem_iter->Prev();
}
});
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
Status s = static_cast_with_check<DBImpl>(db_)->TEST_FlushMemTable();
ASSERT_NOK(s);
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
}
TEST_F(CorruptionTest, DisableKeyOrderCheck) {
ASSERT_OK(db_->SetOptions({{"check_flush_compaction_key_order", "false"}}));
DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
SyncPoint::GetInstance()->SetCallBack(
"OutputValidator::Add:order_check",
[&](void* /*arg*/) { ASSERT_TRUE(false); });
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
ASSERT_OK(db_->Put(WriteOptions(), "foo1", "v1"));
ASSERT_OK(db_->Put(WriteOptions(), "foo3", "v1"));
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(db_->Put(WriteOptions(), "foo2", "v1"));
ASSERT_OK(db_->Put(WriteOptions(), "foo4", "v1"));
ASSERT_OK(dbi->TEST_FlushMemTable());
ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true));
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
}
TEST_F(CorruptionTest, VerifyWholeTableChecksum) {
CloseDb();
Options options;
options.env = env_;
ASSERT_OK(DestroyDB(dbname_, options));
options.create_if_missing = true;
options.file_checksum_gen_factory =
ROCKSDB_NAMESPACE::GetFileChecksumGenCrc32cFactory();
Reopen(&options);
Build(10, 5);
ASSERT_OK(db_->VerifyFileChecksums(ReadOptions()));
CloseDb();
// Corrupt the first byte of each table file, this must be data block.
Corrupt(kTableFile, 0, 1);
ASSERT_OK(TryReopen(&options));
SyncPoint::GetInstance()->DisableProcessing();
SyncPoint::GetInstance()->ClearAllCallBacks();
int count{0};
SyncPoint::GetInstance()->SetCallBack(
"DBImpl::VerifyFullFileChecksum:mismatch", [&](void* arg) {
auto* s = reinterpret_cast<Status*>(arg);
ASSERT_NE(s, nullptr);
++count;
ASSERT_NOK(*s);
});
SyncPoint::GetInstance()->EnableProcessing();
ASSERT_TRUE(db_->VerifyFileChecksums(ReadOptions()).IsCorruption());
ASSERT_EQ(1, count);
}
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
class CrashDuringRecoveryWithCorruptionTest
: public CorruptionTest,
public testing::WithParamInterface<std::tuple<bool, bool>> {
public:
explicit CrashDuringRecoveryWithCorruptionTest()
: CorruptionTest(),
avoid_flush_during_recovery_(std::get<0>(GetParam())),
track_and_verify_wals_in_manifest_(std::get<1>(GetParam())) {}
protected:
const bool avoid_flush_during_recovery_;
const bool track_and_verify_wals_in_manifest_;
};
INSTANTIATE_TEST_CASE_P(CorruptionTest, CrashDuringRecoveryWithCorruptionTest,
::testing::Values(std::make_tuple(true, false),
std::make_tuple(false, false),
std::make_tuple(true, true),
std::make_tuple(false, true)));
// In case of non-TransactionDB with avoid_flush_during_recovery = true, RocksDB
// won't flush the data from WAL to L0 for all column families if possible. As a
// result, not all column families can increase their log_numbers, and
// min_log_number_to_keep won't change.
// It may prematurely persist a new MANIFEST even before we can declare the DB
// is in consistent state after recovery (this is when the new WAL is synced)
// and advances log_numbers for some column families.
//
// If there is power failure before we sync the new WAL, we will end up in
// a situation in which after persisting the MANIFEST, RocksDB will see some
// column families' log_numbers larger than the corrupted wal, and
// "Column family inconsistency: SST file contains data beyond the point of
// corruption" error will be hit, causing recovery to fail.
//
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
// After adding the fix, only after new WAL is synced, RocksDB persist a new
// MANIFEST with column families to ensure RocksDB is in consistent state.
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
// RocksDB writes an empty WriteBatch as a sentinel to the new WAL which is
// synced immediately afterwards. The sequence number of the sentinel
// WriteBatch will be the next sequence number immediately after the largest
// sequence number recovered from previous WALs and MANIFEST because of which DB
// will be in consistent state.
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
// If a future recovery starts from the new MANIFEST, then it means the new WAL
// is successfully synced. Due to the sentinel empty write batch at the
// beginning, kPointInTimeRecovery of WAL is guaranteed to go after this point.
// If future recovery starts from the old MANIFEST, it means the writing the new
// MANIFEST failed. It won't have the "SST ahead of WAL" error.
//
// The combination of corrupting a WAL and injecting an error during subsequent
// re-open exposes the bug of prematurely persisting a new MANIFEST with
// advanced ColumnFamilyData::log_number.
TEST_P(CrashDuringRecoveryWithCorruptionTest, DISABLED_CrashDuringRecovery) {
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
CloseDb();
Options options;
options.track_and_verify_wals_in_manifest =
track_and_verify_wals_in_manifest_;
options.wal_recovery_mode = WALRecoveryMode::kPointInTimeRecovery;
options.avoid_flush_during_recovery = false;
options.env = env_;
ASSERT_OK(DestroyDB(dbname_, options));
options.create_if_missing = true;
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
options.max_write_buffer_number = 8;
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
Reopen(&options);
Status s;
const std::string test_cf_name = "test_cf";
ColumnFamilyHandle* cfh = nullptr;
s = db_->CreateColumnFamily(options, test_cf_name, &cfh);
ASSERT_OK(s);
delete cfh;
CloseDb();
std::vector<ColumnFamilyDescriptor> cf_descs;
cf_descs.emplace_back(kDefaultColumnFamilyName, options);
cf_descs.emplace_back(test_cf_name, options);
std::vector<ColumnFamilyHandle*> handles;
// 1. Open and populate the DB. Write and flush default_cf several times to
// advance wal number so that some column families have advanced log_number
// while other don't.
{
ASSERT_OK(DB::Open(options, dbname_, cf_descs, &handles, &db_));
auto* dbimpl = static_cast_with_check<DBImpl>(db_);
assert(dbimpl);
// Write one key to test_cf.
ASSERT_OK(db_->Put(WriteOptions(), handles[1], "old_key", "dontcare"));
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
ASSERT_OK(db_->Flush(FlushOptions(), handles[1]));
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
// Write to default_cf and flush this cf several times to advance wal
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
// number. TEST_SwitchMemtable makes sure WALs are not synced and test can
// corrupt un-sync WAL.
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
for (int i = 0; i < 2; ++i) {
ASSERT_OK(db_->Put(WriteOptions(), "key" + std::to_string(i), "value"));
ASSERT_OK(dbimpl->TEST_SwitchMemtable());
}
for (auto* h : handles) {
delete h;
}
handles.clear();
CloseDb();
}
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
// 2. Corrupt second last un-syned wal file to emulate power reset which
// caused the DB to lose the un-synced WAL.
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
{
std::vector<uint64_t> file_nums;
GetSortedWalFiles(file_nums);
size_t size = file_nums.size();
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
assert(size >= 2);
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
uint64_t log_num = file_nums[size - 2];
CorruptFileWithTruncation(FileType::kWalFile, log_num,
/*bytes_to_truncate=*/8);
}
// 3. After first crash reopen the DB which contains corrupted WAL. Default
// family has higher log number than corrupted wal number.
//
// Case1: If avoid_flush_during_recovery = true, RocksDB won't flush the data
// from WAL to L0 for all column families (test_cf_name in this case). As a
// result, not all column families can increase their log_numbers, and
// min_log_number_to_keep won't change.
//
// Case2: If avoid_flush_during_recovery = false, all column families have
// flushed their data from WAL to L0 during recovery, and none of them will
// ever need to read the WALs again.
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
// 4. Fault is injected to fail the recovery.
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
{
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
SyncPoint::GetInstance()->DisableProcessing();
SyncPoint::GetInstance()->ClearAllCallBacks();
SyncPoint::GetInstance()->SetCallBack(
"DBImpl::GetLogSizeAndMaybeTruncate:0", [&](void* arg) {
auto* tmp_s = reinterpret_cast<Status*>(arg);
assert(tmp_s);
*tmp_s = Status::IOError("Injected");
});
SyncPoint::GetInstance()->EnableProcessing();
handles.clear();
options.avoid_flush_during_recovery = true;
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
s = DB::Open(options, dbname_, cf_descs, &handles, &db_);
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
ASSERT_TRUE(s.IsIOError());
ASSERT_EQ("IO error: Injected", s.ToString());
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
for (auto* h : handles) {
delete h;
}
CloseDb();
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
SyncPoint::GetInstance()->DisableProcessing();
SyncPoint::GetInstance()->ClearAllCallBacks();
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
}
// 5. After second crash reopen the db with second corruption. Default family
// has higher log number than corrupted wal number.
//
// Case1: If avoid_flush_during_recovery = true, we persist a new
// MANIFEST with advanced log_numbers for some column families only after
// syncing the WAL. So during second crash, RocksDB will skip the corrupted
// WAL files as they have been moved to different folder. Since newly synced
// WAL file's sequence number (sentinel WriteBatch) will be the next
// sequence number immediately after the largest sequence number recovered
// from previous WALs and MANIFEST, db will be in consistent state and opens
// successfully.
//
// Case2: If avoid_flush_during_recovery = false, the corrupted WAL is below
// this number. So during a second crash after persisting the new MANIFEST,
// RocksDB will skip the corrupted WAL(s) because they are all below this
// bound. Therefore, we won't hit the "column family inconsistency" error
// message.
{
options.avoid_flush_during_recovery = avoid_flush_during_recovery_;
ASSERT_OK(DB::Open(options, dbname_, cf_descs, &handles, &db_));
for (auto* h : handles) {
delete h;
}
handles.clear();
CloseDb();
}
}
// In case of TransactionDB, it enables two-phase-commit. The prepare section of
// an uncommitted transaction always need to be kept. Even if we perform flush
// during recovery, we may still need to hold an old WAL. The
// min_log_number_to_keep won't change, and "Column family inconsistency: SST
// file contains data beyond the point of corruption" error will be hit, causing
// recovery to fail.
//
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
// After adding the fix, only after new WAL is synced, RocksDB persist a new
// MANIFEST with column families to ensure RocksDB is in consistent state.
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
// RocksDB writes an empty WriteBatch as a sentinel to the new WAL which is
// synced immediately afterwards. The sequence number of the sentinel
// WriteBatch will be the next sequence number immediately after the largest
// sequence number recovered from previous WALs and MANIFEST because of which DB
// will be in consistent state.
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
// If a future recovery starts from the new MANIFEST, then it means the new WAL
// is successfully synced. Due to the sentinel empty write batch at the
// beginning, kPointInTimeRecovery of WAL is guaranteed to go after this point.
// If future recovery starts from the old MANIFEST, it means the writing the new
// MANIFEST failed. It won't have the "SST ahead of WAL" error.
//
// The combination of corrupting a WAL and injecting an error during subsequent
// re-open exposes the bug of prematurely persisting a new MANIFEST with
// advanced ColumnFamilyData::log_number.
TEST_P(CrashDuringRecoveryWithCorruptionTest,
DISABLED_TxnDbCrashDuringRecovery) {
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
CloseDb();
Options options;
options.wal_recovery_mode = WALRecoveryMode::kPointInTimeRecovery;
options.track_and_verify_wals_in_manifest =
track_and_verify_wals_in_manifest_;
options.avoid_flush_during_recovery = false;
options.env = env_;
ASSERT_OK(DestroyDB(dbname_, options));
options.create_if_missing = true;
options.max_write_buffer_number = 3;
Reopen(&options);
// Create cf test_cf_name.
ColumnFamilyHandle* cfh = nullptr;
const std::string test_cf_name = "test_cf";
Status s = db_->CreateColumnFamily(options, test_cf_name, &cfh);
ASSERT_OK(s);
delete cfh;
CloseDb();
std::vector<ColumnFamilyDescriptor> cf_descs;
cf_descs.emplace_back(kDefaultColumnFamilyName, options);
cf_descs.emplace_back(test_cf_name, options);
std::vector<ColumnFamilyHandle*> handles;
TransactionDB* txn_db = nullptr;
TransactionDBOptions txn_db_opts;
// 1. Open and populate the DB. Write and flush default_cf several times to
// advance wal number so that some column families have advanced log_number
// while other don't.
{
ASSERT_OK(TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs,
&handles, &txn_db));
auto* txn = txn_db->BeginTransaction(WriteOptions(), TransactionOptions());
// Put cf1
ASSERT_OK(txn->Put(handles[1], "foo", "value"));
ASSERT_OK(txn->SetName("txn0"));
ASSERT_OK(txn->Prepare());
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
ASSERT_OK(txn_db->Flush(FlushOptions()));
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
delete txn;
txn = nullptr;
auto* dbimpl = static_cast_with_check<DBImpl>(txn_db->GetRootDB());
assert(dbimpl);
// Put and flush cf0
for (int i = 0; i < 2; ++i) {
ASSERT_OK(txn_db->Put(WriteOptions(), "dontcare", "value"));
ASSERT_OK(dbimpl->TEST_SwitchMemtable());
}
// Put cf1
txn = txn_db->BeginTransaction(WriteOptions(), TransactionOptions());
ASSERT_OK(txn->Put(handles[1], "foo1", "value"));
ASSERT_OK(txn->Commit());
delete txn;
txn = nullptr;
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
for (auto* h : handles) {
delete h;
}
handles.clear();
delete txn_db;
}
// 2. Corrupt second last wal to emulate power reset which caused the DB to
// lose the un-synced WAL.
{
std::vector<uint64_t> file_nums;
GetSortedWalFiles(file_nums);
size_t size = file_nums.size();
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
assert(size >= 2);
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
uint64_t log_num = file_nums[size - 2];
CorruptFileWithTruncation(FileType::kWalFile, log_num,
/*bytes_to_truncate=*/8);
}
// 3. After first crash reopen the DB which contains corrupted WAL. Default
// family has higher log number than corrupted wal number. There may be old
// WAL files that it must not delete because they can contain data of
// uncommitted transactions. As a result, min_log_number_to_keep won't change.
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
{
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
SyncPoint::GetInstance()->DisableProcessing();
SyncPoint::GetInstance()->ClearAllCallBacks();
SyncPoint::GetInstance()->SetCallBack(
"DBImpl::Open::BeforeSyncWAL", [&](void* arg) {
auto* tmp_s = reinterpret_cast<Status*>(arg);
assert(tmp_s);
*tmp_s = Status::IOError("Injected");
});
SyncPoint::GetInstance()->EnableProcessing();
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
handles.clear();
s = TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles,
&txn_db);
ASSERT_TRUE(s.IsIOError());
ASSERT_EQ("IO error: Injected", s.ToString());
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
for (auto* h : handles) {
delete h;
}
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
CloseDb();
SyncPoint::GetInstance()->DisableProcessing();
SyncPoint::GetInstance()->ClearAllCallBacks();
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
}
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
// 4. Corrupt max_wal_num.
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
{
std::vector<uint64_t> file_nums;
GetSortedWalFiles(file_nums);
size_t size = file_nums.size();
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
assert(size >= 2);
Remove corrupted WAL files in kPointRecoveryMode with avoid_flush_duing_recovery set true (#9634) Summary: 1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. 2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail. As a solution, 1. the corrupted WALs whose numbers are larger than the corrupted wal and smaller than the new WAL will be moved to archive folder. 2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634 Test Plan: 1. Added new unit tests 2. make crast_test -j Reviewed By: riversand963 Differential Revision: D34463666 Pulled By: akankshamahajan15 fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
2022-04-12 00:39:31 +02:00
uint64_t log_num = file_nums[size - 1];
CorruptFileWithTruncation(FileType::kWalFile, log_num);
}
// 5. After second crash reopen the db with second corruption. Default family
// has higher log number than corrupted wal number.
// We persist a new MANIFEST with advanced log_numbers for some column
// families only after syncing the WAL. So during second crash, RocksDB will
// skip the corrupted WAL files as they have been moved to different folder.
// Since newly synced WAL file's sequence number (sentinel WriteBatch) will be
// the next sequence number immediately after the largest sequence number
// recovered from previous WALs and MANIFEST, db will be in consistent state
// and opens successfully.
{
ASSERT_OK(TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs,
&handles, &txn_db));
for (auto* h : handles) {
delete h;
}
delete txn_db;
}
}
Update WAL corruption test so that it fails without fix (#9942) Summary: In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't flush the data from WAL to L0 for all column families if possible. As a result, not all column families can increase their log_numbers, and min_log_number_to_keep won't change. For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change. If we persist a new MANIFEST with advanced log_numbers for some column families, then during a second crash after persisting the MANIFEST, RocksDB will see some column families' log_numbers larger than the corrupted WAL, and the "column family inconsistency" error will be hit, causing recovery to fail. This PR update unit tests to emulate the errors and tests are failing without a fix. Error: ``` [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/0, where GetParam() = (true, false) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/1, where GetParam() = (false, false) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/2, where GetParam() = (true, true) (95 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3 db/corruption_test.cc:1190: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF test_cf [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery/3, where GetParam() = (false, true) (92 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/0, where GetParam() = (true, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/1, where GetParam() = (false, false) (97 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/2, where GetParam() = (true, true) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3 db/corruption_test.cc:1354: Failure TransactionDB::Open(options, txn_db_opts, dbname_, cf_descs, &handles, &txn_db) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.TxnDbCrashDuringRecovery/3, where GetParam() = (false, true) (91 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/0, where GetParam() = (true, false) (93 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/1, where GetParam() = (false, false) (94 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/2, where GetParam() = (true, true) (90 ms) [ RUN ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3 db/corruption_test.cc:1483: Failure DB::Open(options, dbname_, cf_descs, &handles, &db_) Corruption: SST file is ahead of WALs in CF default [ FAILED ] CorruptionTest/CrashDuringRecoveryWithCorruptionTest.CrashDuringRecoveryWithFlush/3, where GetParam() = (false, true) (93 ms) [----------] 12 tests from CorruptionTest/CrashDuringRecoveryWithCorruptionTest (1116 ms total) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/9942 Test Plan: Not needed Reviewed By: riversand963 Differential Revision: D36324112 Pulled By: akankshamahajan15 fbshipit-source-id: cab2075ac4ebe48f5ef93a6ea162558aa4fc334d
2022-05-12 01:12:55 +02:00
// This test is similar to
// CrashDuringRecoveryWithCorruptionTest.CrashDuringRecovery except it calls
// flush and corrupts Last WAL. It calls flush to sync some of the WALs and
// remaining are unsyned one of which is then corrupted to simulate crash.
//
// In case of non-TransactionDB with avoid_flush_during_recovery = true, RocksDB
// won't flush the data from WAL to L0 for all column families if possible. As a
// result, not all column families can increase their log_numbers, and
// min_log_number_to_keep won't change.
// It may prematurely persist a new MANIFEST even before we can declare the DB
// is in consistent state after recovery (this is when the new WAL is synced)
// and advances log_numbers for some column families.
//
// If there is power failure before we sync the new WAL, we will end up in
// a situation in which after persisting the MANIFEST, RocksDB will see some
// column families' log_numbers larger than the corrupted wal, and
// "Column family inconsistency: SST file contains data beyond the point of
// corruption" error will be hit, causing recovery to fail.
//
// After adding the fix, only after new WAL is synced, RocksDB persist a new
// MANIFEST with column families to ensure RocksDB is in consistent state.
// RocksDB writes an empty WriteBatch as a sentinel to the new WAL which is
// synced immediately afterwards. The sequence number of the sentinel
// WriteBatch will be the next sequence number immediately after the largest
// sequence number recovered from previous WALs and MANIFEST because of which DB
// will be in consistent state.
// If a future recovery starts from the new MANIFEST, then it means the new WAL
// is successfully synced. Due to the sentinel empty write batch at the
// beginning, kPointInTimeRecovery of WAL is guaranteed to go after this point.
// If future recovery starts from the old MANIFEST, it means the writing the new
// MANIFEST failed. It won't have the "SST ahead of WAL" error.
// The combination of corrupting a WAL and injecting an error during subsequent
// re-open exposes the bug of prematurely persisting a new MANIFEST with
// advanced ColumnFamilyData::log_number.
TEST_P(CrashDuringRecoveryWithCorruptionTest,
DISABLED_CrashDuringRecoveryWithFlush) {
CloseDb();
Options options;
options.wal_recovery_mode = WALRecoveryMode::kPointInTimeRecovery;
options.avoid_flush_during_recovery = false;
options.env = env_;
options.create_if_missing = true;
ASSERT_OK(DestroyDB(dbname_, options));
Reopen(&options);
ColumnFamilyHandle* cfh = nullptr;
const std::string test_cf_name = "test_cf";
Status s = db_->CreateColumnFamily(options, test_cf_name, &cfh);
ASSERT_OK(s);
delete cfh;
CloseDb();
std::vector<ColumnFamilyDescriptor> cf_descs;
cf_descs.emplace_back(kDefaultColumnFamilyName, options);
cf_descs.emplace_back(test_cf_name, options);
std::vector<ColumnFamilyHandle*> handles;
{
ASSERT_OK(DB::Open(options, dbname_, cf_descs, &handles, &db_));
// Write one key to test_cf.
ASSERT_OK(db_->Put(WriteOptions(), handles[1], "old_key", "dontcare"));
// Write to default_cf and flush this cf several times to advance wal
// number.
for (int i = 0; i < 2; ++i) {
ASSERT_OK(db_->Put(WriteOptions(), "key" + std::to_string(i), "value"));
ASSERT_OK(db_->Flush(FlushOptions()));
}
ASSERT_OK(db_->Put(WriteOptions(), handles[1], "dontcare", "dontcare"));
for (auto* h : handles) {
delete h;
}
handles.clear();
CloseDb();
}
// Corrupt second last un-syned wal file to emulate power reset which
// caused the DB to lose the un-synced WAL.
{
std::vector<uint64_t> file_nums;
GetSortedWalFiles(file_nums);
size_t size = file_nums.size();
uint64_t log_num = file_nums[size - 1];
CorruptFileWithTruncation(FileType::kWalFile, log_num,
/*bytes_to_truncate=*/8);
}
// Fault is injected to fail the recovery.
{
SyncPoint::GetInstance()->DisableProcessing();
SyncPoint::GetInstance()->ClearAllCallBacks();
SyncPoint::GetInstance()->SetCallBack(
"DBImpl::GetLogSizeAndMaybeTruncate:0", [&](void* arg) {
auto* tmp_s = reinterpret_cast<Status*>(arg);
assert(tmp_s);
*tmp_s = Status::IOError("Injected");
});
SyncPoint::GetInstance()->EnableProcessing();
handles.clear();
options.avoid_flush_during_recovery = true;
s = DB::Open(options, dbname_, cf_descs, &handles, &db_);
ASSERT_TRUE(s.IsIOError());
ASSERT_EQ("IO error: Injected", s.ToString());
for (auto* h : handles) {
delete h;
}
CloseDb();
SyncPoint::GetInstance()->DisableProcessing();
SyncPoint::GetInstance()->ClearAllCallBacks();
}
// Reopen db again
{
options.avoid_flush_during_recovery = avoid_flush_during_recovery_;
ASSERT_OK(DB::Open(options, dbname_, cf_descs, &handles, &db_));
for (auto* h : handles) {
delete h;
}
}
}
} // namespace ROCKSDB_NAMESPACE
int main(int argc, char** argv) {
ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
::testing::InitGoogleTest(&argc, argv);
RegisterCustomObjects(argc, argv);
rocksdb: switch to gtest Summary: Our existing test notation is very similar to what is used in gtest. It makes it easy to adopt what is different. In this diff I modify existing [[ https://code.google.com/p/googletest/wiki/Primer#Test_Fixtures:_Using_the_Same_Data_Configuration_for_Multiple_Te | test fixture ]] classes to inherit from `testing::Test`. Also for unit tests that use fixture class, `TEST` is replaced with `TEST_F` as required in gtest. There are several custom `main` functions in our existing tests. To make this transition easier, I modify all `main` functions to fallow gtest notation. But eventually we can remove them and use implementation of `main` that gtest provides. ```lang=bash % cat ~/transform #!/bin/sh files=$(git ls-files '*test\.cc') for file in $files do if grep -q "rocksdb::test::RunAllTests()" $file then if grep -Eq '^class \w+Test {' $file then perl -pi -e 's/^(class \w+Test) {/${1}: public testing::Test {/g' $file perl -pi -e 's/^(TEST)/${1}_F/g' $file fi perl -pi -e 's/(int main.*\{)/${1}::testing::InitGoogleTest(&argc, argv);/g' $file perl -pi -e 's/rocksdb::test::RunAllTests/RUN_ALL_TESTS/g' $file fi done % sh ~/transform % make format ``` Second iteration of this diff contains only scripted changes. Third iteration contains manual changes to fix last errors and make it compilable. Test Plan: Build and notice no errors. ```lang=bash % USE_CLANG=1 make check -j55 ``` Tests are still testing. Reviewers: meyering, sdong, rven, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D35157
2015-03-17 22:08:00 +01:00
return RUN_ALL_TESTS();
}
#else
#include <stdio.h>
int main(int /*argc*/, char** /*argv*/) {
fprintf(stderr, "SKIPPED as RepairDB() is not supported in ROCKSDB_LITE\n");
return 0;
}
#endif // !ROCKSDB_LITE