2017-04-06 02:14:05 +02:00
|
|
|
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
|
2017-07-16 01:03:42 +02:00
|
|
|
// 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).
|
2017-04-06 02:14:05 +02:00
|
|
|
//
|
|
|
|
// 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.
|
|
|
|
#include "db/db_impl.h"
|
|
|
|
|
|
|
|
#ifndef __STDC_FORMAT_MACROS
|
|
|
|
#define __STDC_FORMAT_MACROS
|
|
|
|
#endif
|
|
|
|
#include <inttypes.h>
|
2018-05-31 21:53:43 +02:00
|
|
|
#include <set>
|
2018-03-28 19:23:31 +02:00
|
|
|
#include <unordered_set>
|
2017-04-06 02:14:05 +02:00
|
|
|
#include "db/event_helpers.h"
|
Skip deleted WALs during recovery
Summary:
This patch record min log number to keep to the manifest while flushing SST files to ignore them and any WAL older than them during recovery. This is to avoid scenarios when we have a gap between the WAL files are fed to the recovery procedure. The gap could happen by for example out-of-order WAL deletion. Such gap could cause problems in 2PC recovery where the prepared and commit entry are placed into two separate WAL and gap in the WALs could result into not processing the WAL with the commit entry and hence breaking the 2PC recovery logic.
Before the commit, for 2PC case, we determined which log number to keep in FindObsoleteFiles(). We looked at the earliest logs with outstanding prepare entries, or prepare entries whose respective commit or abort are in memtable. With the commit, the same calculation is done while we apply the SST flush. Just before installing the flush file, we precompute the earliest log file to keep after the flush finishes using the same logic (but skipping the memtables just flushed), record this information to the manifest entry for this new flushed SST file. This pre-computed value is also remembered in memory, and will later be used to determine whether a log file can be deleted. This value is unlikely to change until next flush because the commit entry will stay in memtable. (In WritePrepared, we could have removed the older log files as soon as all prepared entries are committed. It's not yet done anyway. Even if we do it, the only thing we loss with this new approach is earlier log deletion between two flushes, which does not guarantee to happen anyway because the obsolete file clean-up function is only executed after flush or compaction)
This min log number to keep is stored in the manifest using the safely-ignore customized field of AddFile entry, in order to guarantee that the DB generated using newer release can be opened by previous releases no older than 4.2.
Closes https://github.com/facebook/rocksdb/pull/3765
Differential Revision: D7747618
Pulled By: siying
fbshipit-source-id: d00c92105b4f83852e9754a1b70d6b64cb590729
2018-05-04 00:35:11 +02:00
|
|
|
#include "db/memtable_list.h"
|
2017-04-06 02:14:05 +02:00
|
|
|
#include "util/file_util.h"
|
|
|
|
#include "util/sst_file_manager_impl.h"
|
|
|
|
|
|
|
|
namespace rocksdb {
|
|
|
|
uint64_t DBImpl::MinLogNumberToKeep() {
|
|
|
|
if (allow_2pc()) {
|
Skip deleted WALs during recovery
Summary:
This patch record min log number to keep to the manifest while flushing SST files to ignore them and any WAL older than them during recovery. This is to avoid scenarios when we have a gap between the WAL files are fed to the recovery procedure. The gap could happen by for example out-of-order WAL deletion. Such gap could cause problems in 2PC recovery where the prepared and commit entry are placed into two separate WAL and gap in the WALs could result into not processing the WAL with the commit entry and hence breaking the 2PC recovery logic.
Before the commit, for 2PC case, we determined which log number to keep in FindObsoleteFiles(). We looked at the earliest logs with outstanding prepare entries, or prepare entries whose respective commit or abort are in memtable. With the commit, the same calculation is done while we apply the SST flush. Just before installing the flush file, we precompute the earliest log file to keep after the flush finishes using the same logic (but skipping the memtables just flushed), record this information to the manifest entry for this new flushed SST file. This pre-computed value is also remembered in memory, and will later be used to determine whether a log file can be deleted. This value is unlikely to change until next flush because the commit entry will stay in memtable. (In WritePrepared, we could have removed the older log files as soon as all prepared entries are committed. It's not yet done anyway. Even if we do it, the only thing we loss with this new approach is earlier log deletion between two flushes, which does not guarantee to happen anyway because the obsolete file clean-up function is only executed after flush or compaction)
This min log number to keep is stored in the manifest using the safely-ignore customized field of AddFile entry, in order to guarantee that the DB generated using newer release can be opened by previous releases no older than 4.2.
Closes https://github.com/facebook/rocksdb/pull/3765
Differential Revision: D7747618
Pulled By: siying
fbshipit-source-id: d00c92105b4f83852e9754a1b70d6b64cb590729
2018-05-04 00:35:11 +02:00
|
|
|
return versions_->min_log_number_to_keep_2pc();
|
|
|
|
} else {
|
|
|
|
return versions_->MinLogNumberWithUnflushedData();
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// * Returns the list of live files in 'sst_live'
|
|
|
|
// If it's doing full scan:
|
|
|
|
// * Returns the list of all files in the filesystem in
|
|
|
|
// 'full_scan_candidate_files'.
|
|
|
|
// Otherwise, gets obsolete files from VersionSet.
|
|
|
|
// no_full_scan = true -- never do the full scan using GetChildren()
|
|
|
|
// force = false -- don't force the full scan, except every
|
|
|
|
// mutable_db_options_.delete_obsolete_files_period_micros
|
|
|
|
// force = true -- force the full scan
|
|
|
|
void DBImpl::FindObsoleteFiles(JobContext* job_context, bool force,
|
|
|
|
bool no_full_scan) {
|
|
|
|
mutex_.AssertHeld();
|
|
|
|
|
|
|
|
// if deletion is disabled, do nothing
|
|
|
|
if (disable_delete_obsolete_files_ > 0) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
bool doing_the_full_scan = false;
|
|
|
|
|
2018-03-08 19:18:34 +01:00
|
|
|
// logic for figuring out if we're doing the full scan
|
2017-04-06 02:14:05 +02:00
|
|
|
if (no_full_scan) {
|
|
|
|
doing_the_full_scan = false;
|
|
|
|
} else if (force ||
|
|
|
|
mutable_db_options_.delete_obsolete_files_period_micros == 0) {
|
|
|
|
doing_the_full_scan = true;
|
|
|
|
} else {
|
|
|
|
const uint64_t now_micros = env_->NowMicros();
|
|
|
|
if ((delete_obsolete_files_last_run_ +
|
|
|
|
mutable_db_options_.delete_obsolete_files_period_micros) <
|
|
|
|
now_micros) {
|
|
|
|
doing_the_full_scan = true;
|
|
|
|
delete_obsolete_files_last_run_ = now_micros;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// don't delete files that might be currently written to from compaction
|
|
|
|
// threads
|
|
|
|
// Since job_context->min_pending_output is set, until file scan finishes,
|
|
|
|
// mutex_ cannot be released. Otherwise, we might see no min_pending_output
|
2018-03-08 19:18:34 +01:00
|
|
|
// here but later find newer generated unfinalized files while scanning.
|
2017-04-06 02:14:05 +02:00
|
|
|
if (!pending_outputs_.empty()) {
|
|
|
|
job_context->min_pending_output = *pending_outputs_.begin();
|
|
|
|
} else {
|
|
|
|
// delete all of them
|
|
|
|
job_context->min_pending_output = std::numeric_limits<uint64_t>::max();
|
|
|
|
}
|
|
|
|
|
|
|
|
// Get obsolete files. This function will also update the list of
|
|
|
|
// pending files in VersionSet().
|
|
|
|
versions_->GetObsoleteFiles(&job_context->sst_delete_files,
|
|
|
|
&job_context->manifest_delete_files,
|
|
|
|
job_context->min_pending_output);
|
|
|
|
|
2018-03-28 19:23:31 +02:00
|
|
|
// Mark the elements in job_context->sst_delete_files as grabbedForPurge
|
|
|
|
// so that other threads calling FindObsoleteFiles with full_scan=true
|
|
|
|
// will not add these files to candidate list for purge.
|
2018-04-06 04:49:06 +02:00
|
|
|
for (const auto& sst_to_del : job_context->sst_delete_files) {
|
|
|
|
MarkAsGrabbedForPurge(sst_to_del.metadata->fd.GetNumber());
|
2018-03-28 19:23:31 +02:00
|
|
|
}
|
|
|
|
|
2017-04-06 02:14:05 +02:00
|
|
|
// store the current filenum, lognum, etc
|
|
|
|
job_context->manifest_file_number = versions_->manifest_file_number();
|
|
|
|
job_context->pending_manifest_file_number =
|
|
|
|
versions_->pending_manifest_file_number();
|
|
|
|
job_context->log_number = MinLogNumberToKeep();
|
|
|
|
job_context->prev_log_number = versions_->prev_log_number();
|
|
|
|
|
|
|
|
versions_->AddLiveFiles(&job_context->sst_live);
|
|
|
|
if (doing_the_full_scan) {
|
2018-03-28 19:23:31 +02:00
|
|
|
InfoLogPrefix info_log_prefix(!immutable_db_options_.db_log_dir.empty(),
|
|
|
|
dbname_);
|
2018-05-31 21:53:43 +02:00
|
|
|
std::set<std::string> paths;
|
2017-04-06 02:14:05 +02:00
|
|
|
for (size_t path_id = 0; path_id < immutable_db_options_.db_paths.size();
|
|
|
|
path_id++) {
|
2018-05-31 21:53:43 +02:00
|
|
|
paths.insert(immutable_db_options_.db_paths[path_id].path);
|
2018-04-06 04:49:06 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
// Note that if cf_paths is not specified in the ColumnFamilyOptions
|
|
|
|
// of a particular column family, we use db_paths as the cf_paths
|
|
|
|
// setting. Hence, there can be multiple duplicates of files from db_paths
|
|
|
|
// in the following code. The duplicate are removed while identifying
|
|
|
|
// unique files in PurgeObsoleteFiles.
|
|
|
|
for (auto cfd : *versions_->GetColumnFamilySet()) {
|
|
|
|
for (size_t path_id = 0; path_id < cfd->ioptions()->cf_paths.size();
|
|
|
|
path_id++) {
|
2018-05-31 21:53:43 +02:00
|
|
|
auto& path = cfd->ioptions()->cf_paths[path_id].path;
|
|
|
|
|
|
|
|
if (paths.find(path) == paths.end()) {
|
|
|
|
paths.insert(path);
|
|
|
|
}
|
2018-04-06 04:49:06 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
for (auto& path : paths) {
|
2017-04-06 02:14:05 +02:00
|
|
|
// set of all files in the directory. We'll exclude files that are still
|
|
|
|
// alive in the subsequent processings.
|
|
|
|
std::vector<std::string> files;
|
2018-04-06 04:49:06 +02:00
|
|
|
env_->GetChildren(path, &files); // Ignore errors
|
2018-03-28 19:23:31 +02:00
|
|
|
for (const std::string& file : files) {
|
|
|
|
uint64_t number;
|
|
|
|
FileType type;
|
|
|
|
// 1. If we cannot parse the file name, we skip;
|
|
|
|
// 2. If the file with file_number equals number has already been
|
|
|
|
// grabbed for purge by another compaction job, or it has already been
|
|
|
|
// schedule for purge, we also skip it if we
|
|
|
|
// are doing full scan in order to avoid double deletion of the same
|
|
|
|
// file under race conditions. See
|
|
|
|
// https://github.com/facebook/rocksdb/issues/3573
|
|
|
|
if (!ParseFileName(file, &number, info_log_prefix.prefix, &type) ||
|
|
|
|
!ShouldPurge(number)) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2017-04-06 02:14:05 +02:00
|
|
|
// TODO(icanadi) clean up this mess to avoid having one-off "/" prefixes
|
|
|
|
job_context->full_scan_candidate_files.emplace_back(
|
2018-04-06 04:49:06 +02:00
|
|
|
"/" + file, path);
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// Add log files in wal_dir
|
|
|
|
if (immutable_db_options_.wal_dir != dbname_) {
|
|
|
|
std::vector<std::string> log_files;
|
|
|
|
env_->GetChildren(immutable_db_options_.wal_dir,
|
|
|
|
&log_files); // Ignore errors
|
2018-03-28 19:23:31 +02:00
|
|
|
for (const std::string& log_file : log_files) {
|
2018-04-06 04:49:06 +02:00
|
|
|
job_context->full_scan_candidate_files.emplace_back(log_file,
|
|
|
|
immutable_db_options_.wal_dir);
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
// Add info log files in db_log_dir
|
|
|
|
if (!immutable_db_options_.db_log_dir.empty() &&
|
|
|
|
immutable_db_options_.db_log_dir != dbname_) {
|
|
|
|
std::vector<std::string> info_log_files;
|
|
|
|
// Ignore errors
|
|
|
|
env_->GetChildren(immutable_db_options_.db_log_dir, &info_log_files);
|
2018-04-06 04:49:06 +02:00
|
|
|
for (std::string& log_file : info_log_files) {
|
|
|
|
job_context->full_scan_candidate_files.emplace_back(log_file,
|
|
|
|
immutable_db_options_.db_log_dir);
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// logs_ is empty when called during recovery, in which case there can't yet
|
|
|
|
// be any tracked obsolete logs
|
|
|
|
if (!alive_log_files_.empty() && !logs_.empty()) {
|
|
|
|
uint64_t min_log_number = job_context->log_number;
|
|
|
|
size_t num_alive_log_files = alive_log_files_.size();
|
|
|
|
// find newly obsoleted log files
|
|
|
|
while (alive_log_files_.begin()->number < min_log_number) {
|
|
|
|
auto& earliest = *alive_log_files_.begin();
|
|
|
|
if (immutable_db_options_.recycle_log_file_num >
|
2018-01-09 21:51:10 +01:00
|
|
|
log_recycle_files_.size()) {
|
2017-04-06 02:14:05 +02:00
|
|
|
ROCKS_LOG_INFO(immutable_db_options_.info_log,
|
|
|
|
"adding log %" PRIu64 " to recycle list\n",
|
|
|
|
earliest.number);
|
2018-01-09 21:51:10 +01:00
|
|
|
log_recycle_files_.push_back(earliest.number);
|
2017-04-06 02:14:05 +02:00
|
|
|
} else {
|
|
|
|
job_context->log_delete_files.push_back(earliest.number);
|
|
|
|
}
|
|
|
|
if (job_context->size_log_to_delete == 0) {
|
|
|
|
job_context->prev_total_log_size = total_log_size_;
|
|
|
|
job_context->num_alive_log_files = num_alive_log_files;
|
|
|
|
}
|
|
|
|
job_context->size_log_to_delete += earliest.size;
|
|
|
|
total_log_size_ -= earliest.size;
|
2017-11-11 02:18:01 +01:00
|
|
|
if (two_write_queues_) {
|
2017-06-30 18:30:03 +02:00
|
|
|
log_write_mutex_.Lock();
|
|
|
|
}
|
|
|
|
alive_log_files_.pop_front();
|
2017-11-11 02:18:01 +01:00
|
|
|
if (two_write_queues_) {
|
2017-06-30 18:30:03 +02:00
|
|
|
log_write_mutex_.Unlock();
|
2017-06-28 22:05:52 +02:00
|
|
|
}
|
2017-04-06 02:14:05 +02:00
|
|
|
// Current log should always stay alive since it can't have
|
|
|
|
// number < MinLogNumber().
|
|
|
|
assert(alive_log_files_.size());
|
|
|
|
}
|
|
|
|
while (!logs_.empty() && logs_.front().number < min_log_number) {
|
|
|
|
auto& log = logs_.front();
|
|
|
|
if (log.getting_synced) {
|
|
|
|
log_sync_cv_.Wait();
|
|
|
|
// logs_ could have changed while we were waiting.
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
logs_to_free_.push_back(log.ReleaseWriter());
|
2017-06-24 23:06:43 +02:00
|
|
|
{
|
|
|
|
InstrumentedMutexLock wl(&log_write_mutex_);
|
|
|
|
logs_.pop_front();
|
|
|
|
}
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
// Current log cannot be obsolete.
|
|
|
|
assert(!logs_.empty());
|
|
|
|
}
|
|
|
|
|
|
|
|
// We're just cleaning up for DB::Write().
|
|
|
|
assert(job_context->logs_to_free.empty());
|
|
|
|
job_context->logs_to_free = logs_to_free_;
|
2018-01-09 21:51:10 +01:00
|
|
|
job_context->log_recycle_files.assign(log_recycle_files_.begin(),
|
|
|
|
log_recycle_files_.end());
|
2018-01-18 02:37:10 +01:00
|
|
|
if (job_context->HaveSomethingToDelete()) {
|
|
|
|
++pending_purge_obsolete_files_;
|
|
|
|
}
|
2017-04-06 02:14:05 +02:00
|
|
|
logs_to_free_.clear();
|
|
|
|
}
|
|
|
|
|
|
|
|
namespace {
|
|
|
|
bool CompareCandidateFile(const JobContext::CandidateFileInfo& first,
|
|
|
|
const JobContext::CandidateFileInfo& second) {
|
|
|
|
if (first.file_name > second.file_name) {
|
|
|
|
return true;
|
|
|
|
} else if (first.file_name < second.file_name) {
|
|
|
|
return false;
|
|
|
|
} else {
|
2018-04-06 04:49:06 +02:00
|
|
|
return (first.file_path > second.file_path);
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}; // namespace
|
|
|
|
|
|
|
|
// Delete obsolete files and log status and information of file deletion
|
2017-09-07 23:11:15 +02:00
|
|
|
void DBImpl::DeleteObsoleteFileImpl(int job_id, const std::string& fname,
|
2018-04-26 22:51:39 +02:00
|
|
|
const std::string& path_to_sync,
|
2018-04-06 04:49:06 +02:00
|
|
|
FileType type, uint64_t number) {
|
2017-09-07 23:11:15 +02:00
|
|
|
Status file_deletion_status;
|
2017-04-06 02:14:05 +02:00
|
|
|
if (type == kTableFile) {
|
|
|
|
file_deletion_status =
|
2018-04-26 22:51:39 +02:00
|
|
|
DeleteSSTFile(&immutable_db_options_, fname, path_to_sync);
|
2017-04-06 02:14:05 +02:00
|
|
|
} else {
|
|
|
|
file_deletion_status = env_->DeleteFile(fname);
|
|
|
|
}
|
2018-03-28 19:23:31 +02:00
|
|
|
TEST_SYNC_POINT_CALLBACK("DBImpl::DeleteObsoleteFileImpl:AfterDeletion",
|
|
|
|
&file_deletion_status);
|
2017-04-06 02:14:05 +02:00
|
|
|
if (file_deletion_status.ok()) {
|
|
|
|
ROCKS_LOG_DEBUG(immutable_db_options_.info_log,
|
|
|
|
"[JOB %d] Delete %s type=%d #%" PRIu64 " -- %s\n", job_id,
|
|
|
|
fname.c_str(), type, number,
|
|
|
|
file_deletion_status.ToString().c_str());
|
|
|
|
} else if (env_->FileExists(fname).IsNotFound()) {
|
|
|
|
ROCKS_LOG_INFO(
|
|
|
|
immutable_db_options_.info_log,
|
|
|
|
"[JOB %d] Tried to delete a non-existing file %s type=%d #%" PRIu64
|
|
|
|
" -- %s\n",
|
|
|
|
job_id, fname.c_str(), type, number,
|
|
|
|
file_deletion_status.ToString().c_str());
|
|
|
|
} else {
|
|
|
|
ROCKS_LOG_ERROR(immutable_db_options_.info_log,
|
|
|
|
"[JOB %d] Failed to delete %s type=%d #%" PRIu64 " -- %s\n",
|
|
|
|
job_id, fname.c_str(), type, number,
|
|
|
|
file_deletion_status.ToString().c_str());
|
|
|
|
}
|
|
|
|
if (type == kTableFile) {
|
|
|
|
EventHelpers::LogAndNotifyTableFileDeletion(
|
|
|
|
&event_logger_, job_id, number, fname, file_deletion_status, GetName(),
|
|
|
|
immutable_db_options_.listeners);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// Diffs the files listed in filenames and those that do not
|
2018-03-08 19:18:34 +01:00
|
|
|
// belong to live files are possibly removed. Also, removes all the
|
2017-04-06 02:14:05 +02:00
|
|
|
// files in sst_delete_files and log_delete_files.
|
|
|
|
// It is not necessary to hold the mutex when invoking this method.
|
2018-04-06 04:49:06 +02:00
|
|
|
void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
|
2018-01-18 02:37:10 +01:00
|
|
|
TEST_SYNC_POINT("DBImpl::PurgeObsoleteFiles:Begin");
|
2017-04-06 02:14:05 +02:00
|
|
|
// we'd better have sth to delete
|
|
|
|
assert(state.HaveSomethingToDelete());
|
|
|
|
|
2018-01-18 02:37:10 +01:00
|
|
|
// FindObsoleteFiles() should've populated this so nonzero
|
|
|
|
assert(state.manifest_file_number != 0);
|
2017-04-06 02:14:05 +02:00
|
|
|
|
|
|
|
// Now, convert live list to an unordered map, WITHOUT mutex held;
|
|
|
|
// set is slow.
|
|
|
|
std::unordered_map<uint64_t, const FileDescriptor*> sst_live_map;
|
|
|
|
for (const FileDescriptor& fd : state.sst_live) {
|
|
|
|
sst_live_map[fd.GetNumber()] = &fd;
|
|
|
|
}
|
|
|
|
std::unordered_set<uint64_t> log_recycle_files_set(
|
|
|
|
state.log_recycle_files.begin(), state.log_recycle_files.end());
|
|
|
|
|
|
|
|
auto candidate_files = state.full_scan_candidate_files;
|
|
|
|
candidate_files.reserve(
|
|
|
|
candidate_files.size() + state.sst_delete_files.size() +
|
|
|
|
state.log_delete_files.size() + state.manifest_delete_files.size());
|
|
|
|
// We may ignore the dbname when generating the file names.
|
|
|
|
const char* kDumbDbName = "";
|
2018-04-06 04:49:06 +02:00
|
|
|
for (auto& file : state.sst_delete_files) {
|
2017-04-06 02:14:05 +02:00
|
|
|
candidate_files.emplace_back(
|
2018-04-06 04:49:06 +02:00
|
|
|
MakeTableFileName(kDumbDbName, file.metadata->fd.GetNumber()), file.path);
|
|
|
|
if (file.metadata->table_reader_handle) {
|
|
|
|
table_cache_->Release(file.metadata->table_reader_handle);
|
2017-07-27 21:10:49 +02:00
|
|
|
}
|
2018-04-06 04:49:06 +02:00
|
|
|
file.DeleteMetadata();
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
for (auto file_num : state.log_delete_files) {
|
|
|
|
if (file_num > 0) {
|
2018-04-06 04:49:06 +02:00
|
|
|
candidate_files.emplace_back(LogFileName(kDumbDbName, file_num),
|
|
|
|
immutable_db_options_.wal_dir);
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
for (const auto& filename : state.manifest_delete_files) {
|
2018-04-06 04:49:06 +02:00
|
|
|
candidate_files.emplace_back(filename, dbname_);
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
// dedup state.candidate_files so we don't try to delete the same
|
|
|
|
// file twice
|
|
|
|
std::sort(candidate_files.begin(), candidate_files.end(),
|
|
|
|
CompareCandidateFile);
|
|
|
|
candidate_files.erase(
|
|
|
|
std::unique(candidate_files.begin(), candidate_files.end()),
|
|
|
|
candidate_files.end());
|
|
|
|
|
|
|
|
if (state.prev_total_log_size > 0) {
|
|
|
|
ROCKS_LOG_INFO(immutable_db_options_.info_log,
|
|
|
|
"[JOB %d] Try to delete WAL files size %" PRIu64
|
|
|
|
", prev total WAL file size %" PRIu64
|
|
|
|
", number of live WAL files %" ROCKSDB_PRIszt ".\n",
|
|
|
|
state.job_id, state.size_log_to_delete,
|
|
|
|
state.prev_total_log_size, state.num_alive_log_files);
|
|
|
|
}
|
|
|
|
|
|
|
|
std::vector<std::string> old_info_log_files;
|
|
|
|
InfoLogPrefix info_log_prefix(!immutable_db_options_.db_log_dir.empty(),
|
|
|
|
dbname_);
|
2018-03-28 19:23:31 +02:00
|
|
|
std::unordered_set<uint64_t> files_to_del;
|
2017-04-06 02:14:05 +02:00
|
|
|
for (const auto& candidate_file : candidate_files) {
|
|
|
|
std::string to_delete = candidate_file.file_name;
|
|
|
|
uint64_t number;
|
|
|
|
FileType type;
|
|
|
|
// Ignore file if we cannot recognize it.
|
|
|
|
if (!ParseFileName(to_delete, &number, info_log_prefix.prefix, &type)) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
bool keep = true;
|
|
|
|
switch (type) {
|
|
|
|
case kLogFile:
|
|
|
|
keep = ((number >= state.log_number) ||
|
|
|
|
(number == state.prev_log_number) ||
|
|
|
|
(log_recycle_files_set.find(number) !=
|
|
|
|
log_recycle_files_set.end()));
|
|
|
|
break;
|
|
|
|
case kDescriptorFile:
|
|
|
|
// Keep my manifest file, and any newer incarnations'
|
|
|
|
// (can happen during manifest roll)
|
|
|
|
keep = (number >= state.manifest_file_number);
|
|
|
|
break;
|
|
|
|
case kTableFile:
|
|
|
|
// If the second condition is not there, this makes
|
|
|
|
// DontDeletePendingOutputs fail
|
|
|
|
keep = (sst_live_map.find(number) != sst_live_map.end()) ||
|
|
|
|
number >= state.min_pending_output;
|
2018-03-28 19:23:31 +02:00
|
|
|
if (!keep) {
|
|
|
|
files_to_del.insert(number);
|
|
|
|
}
|
2017-04-06 02:14:05 +02:00
|
|
|
break;
|
|
|
|
case kTempFile:
|
|
|
|
// Any temp files that are currently being written to must
|
|
|
|
// be recorded in pending_outputs_, which is inserted into "live".
|
|
|
|
// Also, SetCurrentFile creates a temp file when writing out new
|
|
|
|
// manifest, which is equal to state.pending_manifest_file_number. We
|
|
|
|
// should not delete that file
|
|
|
|
//
|
|
|
|
// TODO(yhchiang): carefully modify the third condition to safely
|
|
|
|
// remove the temp options files.
|
|
|
|
keep = (sst_live_map.find(number) != sst_live_map.end()) ||
|
|
|
|
(number == state.pending_manifest_file_number) ||
|
|
|
|
(to_delete.find(kOptionsFileNamePrefix) != std::string::npos);
|
|
|
|
break;
|
|
|
|
case kInfoLogFile:
|
|
|
|
keep = true;
|
|
|
|
if (number != 0) {
|
|
|
|
old_info_log_files.push_back(to_delete);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case kCurrentFile:
|
|
|
|
case kDBLockFile:
|
|
|
|
case kIdentityFile:
|
|
|
|
case kMetaDatabase:
|
|
|
|
case kOptionsFile:
|
2017-04-18 21:00:36 +02:00
|
|
|
case kBlobFile:
|
2017-04-06 02:14:05 +02:00
|
|
|
keep = true;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (keep) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
std::string fname;
|
2018-04-26 22:51:39 +02:00
|
|
|
std::string dir_to_sync;
|
2017-04-06 02:14:05 +02:00
|
|
|
if (type == kTableFile) {
|
|
|
|
// evict from cache
|
|
|
|
TableCache::Evict(table_cache_.get(), number);
|
2018-04-06 04:49:06 +02:00
|
|
|
fname = MakeTableFileName(candidate_file.file_path, number);
|
2018-04-26 22:51:39 +02:00
|
|
|
dir_to_sync = candidate_file.file_path;
|
2017-04-06 02:14:05 +02:00
|
|
|
} else {
|
2018-04-26 22:51:39 +02:00
|
|
|
dir_to_sync =
|
|
|
|
(type == kLogFile) ? immutable_db_options_.wal_dir : dbname_;
|
|
|
|
fname = dir_to_sync + "/" + to_delete;
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
#ifndef ROCKSDB_LITE
|
|
|
|
if (type == kLogFile && (immutable_db_options_.wal_ttl_seconds > 0 ||
|
|
|
|
immutable_db_options_.wal_size_limit_mb > 0)) {
|
|
|
|
wal_manager_.ArchiveWALFile(fname, number);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
#endif // !ROCKSDB_LITE
|
|
|
|
|
|
|
|
Status file_deletion_status;
|
|
|
|
if (schedule_only) {
|
|
|
|
InstrumentedMutexLock guard_lock(&mutex_);
|
2018-04-26 22:51:39 +02:00
|
|
|
SchedulePendingPurge(fname, dir_to_sync, type, number, state.job_id);
|
2017-04-06 02:14:05 +02:00
|
|
|
} else {
|
2018-04-26 22:51:39 +02:00
|
|
|
DeleteObsoleteFileImpl(state.job_id, fname, dir_to_sync, type, number);
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-03-28 19:23:31 +02:00
|
|
|
{
|
|
|
|
// After purging obsolete files, remove them from files_grabbed_for_purge_.
|
|
|
|
// Use a temporary vector to perform bulk deletion via swap.
|
|
|
|
InstrumentedMutexLock guard_lock(&mutex_);
|
|
|
|
std::vector<uint64_t> tmp;
|
|
|
|
for (auto fn : files_grabbed_for_purge_) {
|
|
|
|
if (files_to_del.count(fn) == 0) {
|
|
|
|
tmp.emplace_back(fn);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
files_grabbed_for_purge_.swap(tmp);
|
|
|
|
}
|
|
|
|
|
2017-04-06 02:14:05 +02:00
|
|
|
// Delete old info log files.
|
|
|
|
size_t old_info_log_file_count = old_info_log_files.size();
|
|
|
|
if (old_info_log_file_count != 0 &&
|
|
|
|
old_info_log_file_count >= immutable_db_options_.keep_log_file_num) {
|
|
|
|
std::sort(old_info_log_files.begin(), old_info_log_files.end());
|
|
|
|
size_t end =
|
|
|
|
old_info_log_file_count - immutable_db_options_.keep_log_file_num;
|
|
|
|
for (unsigned int i = 0; i <= end; i++) {
|
|
|
|
std::string& to_delete = old_info_log_files.at(i);
|
|
|
|
std::string full_path_to_delete =
|
|
|
|
(immutable_db_options_.db_log_dir.empty()
|
|
|
|
? dbname_
|
|
|
|
: immutable_db_options_.db_log_dir) +
|
|
|
|
"/" + to_delete;
|
|
|
|
ROCKS_LOG_INFO(immutable_db_options_.info_log,
|
|
|
|
"[JOB %d] Delete info log file %s\n", state.job_id,
|
|
|
|
full_path_to_delete.c_str());
|
|
|
|
Status s = env_->DeleteFile(full_path_to_delete);
|
|
|
|
if (!s.ok()) {
|
|
|
|
if (env_->FileExists(full_path_to_delete).IsNotFound()) {
|
|
|
|
ROCKS_LOG_INFO(
|
|
|
|
immutable_db_options_.info_log,
|
|
|
|
"[JOB %d] Tried to delete non-existing info log file %s FAILED "
|
|
|
|
"-- %s\n",
|
|
|
|
state.job_id, to_delete.c_str(), s.ToString().c_str());
|
|
|
|
} else {
|
|
|
|
ROCKS_LOG_ERROR(immutable_db_options_.info_log,
|
|
|
|
"[JOB %d] Delete info log file %s FAILED -- %s\n",
|
|
|
|
state.job_id, to_delete.c_str(),
|
|
|
|
s.ToString().c_str());
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#ifndef ROCKSDB_LITE
|
|
|
|
wal_manager_.PurgeObsoleteWALFiles();
|
|
|
|
#endif // ROCKSDB_LITE
|
|
|
|
LogFlush(immutable_db_options_.info_log);
|
2018-01-18 02:37:10 +01:00
|
|
|
InstrumentedMutexLock l(&mutex_);
|
|
|
|
--pending_purge_obsolete_files_;
|
|
|
|
assert(pending_purge_obsolete_files_ >= 0);
|
|
|
|
if (pending_purge_obsolete_files_ == 0) {
|
|
|
|
bg_cv_.SignalAll();
|
|
|
|
}
|
|
|
|
TEST_SYNC_POINT("DBImpl::PurgeObsoleteFiles:End");
|
2017-04-06 02:14:05 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
void DBImpl::DeleteObsoleteFiles() {
|
|
|
|
mutex_.AssertHeld();
|
|
|
|
JobContext job_context(next_job_id_.fetch_add(1));
|
|
|
|
FindObsoleteFiles(&job_context, true);
|
|
|
|
|
|
|
|
mutex_.Unlock();
|
|
|
|
if (job_context.HaveSomethingToDelete()) {
|
|
|
|
PurgeObsoleteFiles(job_context);
|
|
|
|
}
|
|
|
|
job_context.Clean();
|
|
|
|
mutex_.Lock();
|
|
|
|
}
|
2018-03-28 19:23:31 +02:00
|
|
|
|
Skip deleted WALs during recovery
Summary:
This patch record min log number to keep to the manifest while flushing SST files to ignore them and any WAL older than them during recovery. This is to avoid scenarios when we have a gap between the WAL files are fed to the recovery procedure. The gap could happen by for example out-of-order WAL deletion. Such gap could cause problems in 2PC recovery where the prepared and commit entry are placed into two separate WAL and gap in the WALs could result into not processing the WAL with the commit entry and hence breaking the 2PC recovery logic.
Before the commit, for 2PC case, we determined which log number to keep in FindObsoleteFiles(). We looked at the earliest logs with outstanding prepare entries, or prepare entries whose respective commit or abort are in memtable. With the commit, the same calculation is done while we apply the SST flush. Just before installing the flush file, we precompute the earliest log file to keep after the flush finishes using the same logic (but skipping the memtables just flushed), record this information to the manifest entry for this new flushed SST file. This pre-computed value is also remembered in memory, and will later be used to determine whether a log file can be deleted. This value is unlikely to change until next flush because the commit entry will stay in memtable. (In WritePrepared, we could have removed the older log files as soon as all prepared entries are committed. It's not yet done anyway. Even if we do it, the only thing we loss with this new approach is earlier log deletion between two flushes, which does not guarantee to happen anyway because the obsolete file clean-up function is only executed after flush or compaction)
This min log number to keep is stored in the manifest using the safely-ignore customized field of AddFile entry, in order to guarantee that the DB generated using newer release can be opened by previous releases no older than 4.2.
Closes https://github.com/facebook/rocksdb/pull/3765
Differential Revision: D7747618
Pulled By: siying
fbshipit-source-id: d00c92105b4f83852e9754a1b70d6b64cb590729
2018-05-04 00:35:11 +02:00
|
|
|
uint64_t FindMinPrepLogReferencedByMemTable(
|
|
|
|
VersionSet* vset, const ColumnFamilyData* cfd_to_flush,
|
|
|
|
const autovector<MemTable*>& memtables_to_flush) {
|
|
|
|
uint64_t min_log = 0;
|
|
|
|
|
|
|
|
// we must look through the memtables for two phase transactions
|
|
|
|
// that have been committed but not yet flushed
|
|
|
|
for (auto loop_cfd : *vset->GetColumnFamilySet()) {
|
|
|
|
if (loop_cfd->IsDropped() || loop_cfd == cfd_to_flush) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
auto log = loop_cfd->imm()->PrecomputeMinLogContainingPrepSection(
|
|
|
|
memtables_to_flush);
|
|
|
|
|
|
|
|
if (log > 0 && (min_log == 0 || log < min_log)) {
|
|
|
|
min_log = log;
|
|
|
|
}
|
|
|
|
|
|
|
|
log = loop_cfd->mem()->GetMinLogContainingPrepSection();
|
|
|
|
|
|
|
|
if (log > 0 && (min_log == 0 || log < min_log)) {
|
|
|
|
min_log = log;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return min_log;
|
|
|
|
}
|
|
|
|
|
|
|
|
uint64_t PrecomputeMinLogNumberToKeep(
|
|
|
|
VersionSet* vset, const ColumnFamilyData& cfd_to_flush,
|
|
|
|
autovector<VersionEdit*> edit_list,
|
|
|
|
const autovector<MemTable*>& memtables_to_flush,
|
|
|
|
LogsWithPrepTracker* prep_tracker) {
|
|
|
|
assert(vset != nullptr);
|
|
|
|
assert(prep_tracker != nullptr);
|
|
|
|
// Calculate updated min_log_number_to_keep
|
|
|
|
// Since the function should only be called in 2pc mode, log number in
|
|
|
|
// the version edit should be sufficient.
|
|
|
|
|
|
|
|
// Precompute the min log number containing unflushed data for the column
|
|
|
|
// family being flushed (`cfd_to_flush`).
|
|
|
|
uint64_t cf_min_log_number_to_keep = 0;
|
|
|
|
for (auto& e : edit_list) {
|
|
|
|
if (e->has_log_number()) {
|
|
|
|
cf_min_log_number_to_keep =
|
|
|
|
std::max(cf_min_log_number_to_keep, e->log_number());
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (cf_min_log_number_to_keep == 0) {
|
|
|
|
// No version edit contains information on log number. The log number
|
|
|
|
// for this column family should stay the same as it is.
|
|
|
|
cf_min_log_number_to_keep = cfd_to_flush.GetLogNumber();
|
|
|
|
}
|
|
|
|
|
|
|
|
// Get min log number containing unflushed data for other column families.
|
|
|
|
uint64_t min_log_number_to_keep =
|
|
|
|
vset->PreComputeMinLogNumberWithUnflushedData(&cfd_to_flush);
|
|
|
|
if (cf_min_log_number_to_keep != 0) {
|
|
|
|
min_log_number_to_keep =
|
|
|
|
std::min(cf_min_log_number_to_keep, min_log_number_to_keep);
|
|
|
|
}
|
|
|
|
|
|
|
|
// if are 2pc we must consider logs containing prepared
|
|
|
|
// sections of outstanding transactions.
|
|
|
|
//
|
|
|
|
// We must check min logs with outstanding prep before we check
|
|
|
|
// logs references by memtables because a log referenced by the
|
|
|
|
// first data structure could transition to the second under us.
|
|
|
|
//
|
|
|
|
// TODO: iterating over all column families under db mutex.
|
|
|
|
// should find more optimal solution
|
|
|
|
auto min_log_in_prep_heap =
|
|
|
|
prep_tracker->FindMinLogContainingOutstandingPrep();
|
|
|
|
|
|
|
|
if (min_log_in_prep_heap != 0 &&
|
|
|
|
min_log_in_prep_heap < min_log_number_to_keep) {
|
|
|
|
min_log_number_to_keep = min_log_in_prep_heap;
|
|
|
|
}
|
|
|
|
|
|
|
|
uint64_t min_log_refed_by_mem = FindMinPrepLogReferencedByMemTable(
|
|
|
|
vset, &cfd_to_flush, memtables_to_flush);
|
|
|
|
|
|
|
|
if (min_log_refed_by_mem != 0 &&
|
|
|
|
min_log_refed_by_mem < min_log_number_to_keep) {
|
|
|
|
min_log_number_to_keep = min_log_refed_by_mem;
|
|
|
|
}
|
|
|
|
return min_log_number_to_keep;
|
|
|
|
}
|
|
|
|
|
2017-04-06 02:14:05 +02:00
|
|
|
} // namespace rocksdb
|