rocksdb/utilities/fault_injection_env.h

253 lines
8.0 KiB
C
Raw Permalink 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 2014 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.
// This test uses a custom Env to keep track of the state of a filesystem as of
// the last "sync". It then checks for data loss errors by purposely dropping
// file data (or entire files) not protected by a "sync".
#pragma once
#include <map>
#include <set>
#include <string>
#include "file/filename.h"
#include "rocksdb/env.h"
#include "util/mutexlock.h"
namespace ROCKSDB_NAMESPACE {
class Random;
class TestWritableFile;
class FaultInjectionTestEnv;
struct FileState {
std::string filename_;
ssize_t pos_;
ssize_t pos_at_last_sync_;
ssize_t pos_at_last_flush_;
explicit FileState(const std::string& filename)
: filename_(filename),
pos_(-1),
pos_at_last_sync_(-1),
pos_at_last_flush_(-1) {}
FileState() : pos_(-1), pos_at_last_sync_(-1), pos_at_last_flush_(-1) {}
bool IsFullySynced() const { return pos_ <= 0 || pos_ == pos_at_last_sync_; }
Status DropUnsyncedData(Env* env) const;
Status DropRandomUnsyncedData(Env* env, Random* rand) const;
};
class TestRandomAccessFile : public RandomAccessFile {
public:
TestRandomAccessFile(std::unique_ptr<RandomAccessFile>&& target,
FaultInjectionTestEnv* env);
Status Read(uint64_t offset, size_t n, Slice* result,
char* scratch) const override;
Status Prefetch(uint64_t offset, size_t n) override;
Status MultiRead(ReadRequest* reqs, size_t num_reqs) override;
private:
std::unique_ptr<RandomAccessFile> target_;
FaultInjectionTestEnv* env_;
};
// A wrapper around WritableFileWriter* file
// is written to or sync'ed.
class TestWritableFile : public WritableFile {
public:
explicit TestWritableFile(const std::string& fname,
std::unique_ptr<WritableFile>&& f,
FaultInjectionTestEnv* env);
virtual ~TestWritableFile();
virtual Status Append(const Slice& data) override;
virtual Status Append(
const Slice& data,
const DataVerificationInfo& /*verification_info*/) override {
return Append(data);
}
virtual Status Truncate(uint64_t size) override {
return target_->Truncate(size);
}
virtual Status Close() override;
virtual Status Flush() override;
virtual Status Sync() override;
virtual bool IsSyncThreadSafe() const override { return true; }
virtual Status PositionedAppend(const Slice& data,
uint64_t offset) override {
return target_->PositionedAppend(data, offset);
}
virtual Status PositionedAppend(
const Slice& data, uint64_t offset,
const DataVerificationInfo& /*verification_info*/) override {
return PositionedAppend(data, offset);
}
virtual bool use_direct_io() const override {
return target_->use_direct_io();
};
private:
FileState state_;
std::unique_ptr<WritableFile> target_;
bool writable_file_opened_;
FaultInjectionTestEnv* env_;
};
// A wrapper around WritableFileWriter* file
// is written to or sync'ed.
class TestRandomRWFile : public RandomRWFile {
public:
explicit TestRandomRWFile(const std::string& fname,
std::unique_ptr<RandomRWFile>&& f,
FaultInjectionTestEnv* env);
virtual ~TestRandomRWFile();
Status Write(uint64_t offset, const Slice& data) override;
Status Read(uint64_t offset, size_t n, Slice* result,
char* scratch) const override;
Status Close() override;
Status Flush() override;
Status Sync() override;
size_t GetRequiredBufferAlignment() const override {
return target_->GetRequiredBufferAlignment();
}
bool use_direct_io() const override { return target_->use_direct_io(); };
private:
std::unique_ptr<RandomRWFile> target_;
bool file_opened_;
FaultInjectionTestEnv* env_;
};
class TestDirectory : public Directory {
public:
explicit TestDirectory(FaultInjectionTestEnv* env, std::string dirname,
Directory* dir)
: env_(env), dirname_(dirname), dir_(dir) {}
~TestDirectory() {}
virtual Status Fsync() override;
private:
FaultInjectionTestEnv* env_;
std::string dirname_;
std::unique_ptr<Directory> dir_;
};
class FaultInjectionTestEnv : public EnvWrapper {
public:
explicit FaultInjectionTestEnv(Env* base)
: EnvWrapper(base), filesystem_active_(true) {}
virtual ~FaultInjectionTestEnv() { error_.PermitUncheckedError(); }
Status NewDirectory(const std::string& name,
std::unique_ptr<Directory>* result) override;
Status NewWritableFile(const std::string& fname,
std::unique_ptr<WritableFile>* result,
const EnvOptions& soptions) override;
Status ReopenWritableFile(const std::string& fname,
std::unique_ptr<WritableFile>* result,
const EnvOptions& soptions) override;
Status NewRandomRWFile(const std::string& fname,
std::unique_ptr<RandomRWFile>* result,
const EnvOptions& soptions) override;
Status NewRandomAccessFile(const std::string& fname,
std::unique_ptr<RandomAccessFile>* result,
const EnvOptions& soptions) override;
virtual Status DeleteFile(const std::string& f) override;
virtual Status RenameFile(const std::string& s,
const std::string& t) override;
Introduce a new MultiGet batching implementation (#5011) Summary: This PR introduces a new MultiGet() API, with the underlying implementation grouping keys based on SST file and batching lookups in a file. The reason for the new API is twofold - the definition allows callers to allocate storage for status and values on stack instead of std::vector, as well as return values as PinnableSlices in order to avoid copying, and it keeps the original MultiGet() implementation intact while we experiment with batching. Batching is useful when there is some spatial locality to the keys being queries, as well as larger batch sizes. The main benefits are due to - 1. Fewer function calls, especially to BlockBasedTableReader::MultiGet() and FullFilterBlockReader::KeysMayMatch() 2. Bloom filter cachelines can be prefetched, hiding the cache miss latency The next step is to optimize the binary searches in the level_storage_info, index blocks and data blocks, since we could reduce the number of key comparisons if the keys are relatively close to each other. The batching optimizations also need to be extended to other formats, such as PlainTable and filter formats. This also needs to be added to db_stress. Benchmark results from db_bench for various batch size/locality of reference combinations are given below. Locality was simulated by offsetting the keys in a batch by a stride length. Each SST file is about 8.6MB uncompressed and key/value size is 16/100 uncompressed. To focus on the cpu benefit of batching, the runs were single threaded and bound to the same cpu to eliminate interference from other system events. The results show a 10-25% improvement in micros/op from smaller to larger batch sizes (4 - 32). Batch Sizes 1 | 2 | 4 | 8 | 16 | 32 Random pattern (Stride length 0) 4.158 | 4.109 | 4.026 | 4.05 | 4.1 | 4.074 - Get 4.438 | 4.302 | 4.165 | 4.122 | 4.096 | 4.075 - MultiGet (no batching) 4.461 | 4.256 | 4.277 | 4.11 | 4.182 | 4.14 - MultiGet (w/ batching) Good locality (Stride length 16) 4.048 | 3.659 | 3.248 | 2.99 | 2.84 | 2.753 4.429 | 3.728 | 3.406 | 3.053 | 2.911 | 2.781 4.452 | 3.45 | 2.833 | 2.451 | 2.233 | 2.135 Good locality (Stride length 256) 4.066 | 3.786 | 3.581 | 3.447 | 3.415 | 3.232 4.406 | 4.005 | 3.644 | 3.49 | 3.381 | 3.268 4.393 | 3.649 | 3.186 | 2.882 | 2.676 | 2.62 Medium locality (Stride length 4096) 4.012 | 3.922 | 3.768 | 3.61 | 3.582 | 3.555 4.364 | 4.057 | 3.791 | 3.65 | 3.57 | 3.465 4.479 | 3.758 | 3.316 | 3.077 | 2.959 | 2.891 dbbench command used (on a DB with 4 levels, 12 million keys)- TEST_TMPDIR=/dev/shm numactl -C 10 ./db_bench.tmp -use_existing_db=true -benchmarks="readseq,multireadrandom" -write_buffer_size=4194304 -target_file_size_base=4194304 -max_bytes_for_level_base=16777216 -num=12000000 -reads=12000000 -duration=90 -threads=1 -compression_type=none -cache_size=4194304000 -batch_size=32 -disable_auto_compactions=true -bloom_bits=10 -cache_index_and_filter_blocks=true -pin_l0_filter_and_index_blocks_in_cache=true -multiread_batched=true -multiread_stride=4 Pull Request resolved: https://github.com/facebook/rocksdb/pull/5011 Differential Revision: D14348703 Pulled By: anand1976 fbshipit-source-id: 774406dab3776d979c809522a67bedac6c17f84b
2019-04-11 23:24:09 +02:00
// Undef to eliminate clash on Windows
#undef GetFreeSpace
Auto recovery from out of space errors (#4164) Summary: This commit implements automatic recovery from a Status::NoSpace() error during background operations such as write callback, flush and compaction. The broad design is as follows - 1. Compaction errors are treated as soft errors and don't put the database in read-only mode. A compaction is delayed until enough free disk space is available to accomodate the compaction outputs, which is estimated based on the input size. This means that users can continue to write, and we rely on the WriteController to delay or stop writes if the compaction debt becomes too high due to persistent low disk space condition 2. Errors during write callback and flush are treated as hard errors, i.e the database is put in read-only mode and goes back to read-write only fater certain recovery actions are taken. 3. Both types of recovery rely on the SstFileManagerImpl to poll for sufficient disk space. We assume that there is a 1-1 mapping between an SFM and the underlying OS storage container. For cases where multiple DBs are hosted on a single storage container, the user is expected to allocate a single SFM instance and use the same one for all the DBs. If no SFM is specified by the user, DBImpl::Open() will allocate one, but this will be one per DB and each DB will recover independently. The recovery implemented by SFM is as follows - a) On the first occurance of an out of space error during compaction, subsequent compactions will be delayed until the disk free space check indicates enough available space. The required space is computed as the sum of input sizes. b) The free space check requirement will be removed once the amount of free space is greater than the size reserved by in progress compactions when the first error occured c) If the out of space error is a hard error, a background thread in SFM will poll for sufficient headroom before triggering the recovery of the database and putting it in write-only mode. The headroom is calculated as the sum of the write_buffer_size of all the DB instances associated with the SFM 4. EventListener callbacks will be called at the start and completion of automatic recovery. Users can disable the auto recov ery in the start callback, and later initiate it manually by calling DB::Resume() Todo: 1. More extensive testing 2. Add disk full condition to db_stress (follow-on PR) Pull Request resolved: https://github.com/facebook/rocksdb/pull/4164 Differential Revision: D9846378 Pulled By: anand1976 fbshipit-source-id: 80ea875dbd7f00205e19c82215ff6e37da10da4a
2018-09-15 22:36:19 +02:00
virtual Status GetFreeSpace(const std::string& path,
uint64_t* disk_free) override {
if (!IsFilesystemActive() && error_ == Status::NoSpace()) {
*disk_free = 0;
return Status::OK();
} else {
return target()->GetFreeSpace(path, disk_free);
}
}
void WritableFileClosed(const FileState& state);
void WritableFileSynced(const FileState& state);
void WritableFileAppended(const FileState& state);
// For every file that is not fully synced, make a call to `func` with
// FileState of the file as the parameter.
Status DropFileData(std::function<Status(Env*, FileState)> func);
Status DropUnsyncedFileData();
Status DropRandomUnsyncedFileData(Random* rnd);
Status DeleteFilesCreatedAfterLastDirSync();
void ResetState();
void UntrackFile(const std::string& f);
void SyncDir(const std::string& dirname) {
MutexLock l(&mutex_);
dir_to_new_files_since_last_sync_.erase(dirname);
}
// Setting the filesystem to inactive is the test equivalent to simulating a
// system reset. Setting to inactive will freeze our saved filesystem state so
// that it will stop being recorded. It can then be reset back to the state at
// the time of the reset.
bool IsFilesystemActive() {
MutexLock l(&mutex_);
return filesystem_active_;
}
void SetFilesystemActiveNoLock(bool active,
Status error = Status::Corruption("Not active")) {
error.PermitUncheckedError();
filesystem_active_ = active;
if (!active) {
error_ = error;
}
error.PermitUncheckedError();
}
void SetFilesystemActive(bool active,
Status error = Status::Corruption("Not active")) {
error.PermitUncheckedError();
MutexLock l(&mutex_);
SetFilesystemActiveNoLock(active, error);
error.PermitUncheckedError();
}
void AssertNoOpenFile() { assert(open_files_.empty()); }
Status GetError() { return error_; }
private:
port::Mutex mutex_;
std::map<std::string, FileState> db_file_state_;
std::set<std::string> open_files_;
std::unordered_map<std::string, std::set<std::string>>
dir_to_new_files_since_last_sync_;
bool filesystem_active_; // Record flushes, syncs, writes
Status error_;
};
} // namespace ROCKSDB_NAMESPACE