You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
rocksdb/db/db_impl/db_impl_secondary.cc

675 lines
25 KiB

Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
// 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).
#include "db/db_impl/db_impl_secondary.h"
#include <cinttypes>
#include "db/arena_wrapped_db_iter.h"
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
#include "db/merge_context.h"
#include "logging/auto_roll_logger.h"
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
#include "monitoring/perf_context_imp.h"
#include "util/cast_util.h"
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
namespace ROCKSDB_NAMESPACE {
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
#ifndef ROCKSDB_LITE
DBImplSecondary::DBImplSecondary(const DBOptions& db_options,
const std::string& dbname)
: DBImpl(db_options, dbname) {
ROCKS_LOG_INFO(immutable_db_options_.info_log,
"Opening the db in secondary mode");
LogFlush(immutable_db_options_.info_log);
}
DBImplSecondary::~DBImplSecondary() {}
Status DBImplSecondary::Recover(
const std::vector<ColumnFamilyDescriptor>& column_families,
bool /*readonly*/, bool /*error_if_log_file_exist*/,
bool /*error_if_data_exists_in_logs*/, uint64_t*) {
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
mutex_.AssertHeld();
JobContext job_context(0);
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
Status s;
s = static_cast<ReactiveVersionSet*>(versions_.get())
->Recover(column_families, &manifest_reader_, &manifest_reporter_,
&manifest_reader_status_);
if (!s.ok()) {
return s;
}
if (immutable_db_options_.paranoid_checks && s.ok()) {
s = CheckConsistency();
}
// Initial max_total_in_memory_state_ before recovery logs.
max_total_in_memory_state_ = 0;
for (auto cfd : *versions_->GetColumnFamilySet()) {
auto* mutable_cf_options = cfd->GetLatestMutableCFOptions();
max_total_in_memory_state_ += mutable_cf_options->write_buffer_size *
mutable_cf_options->max_write_buffer_number;
}
if (s.ok()) {
default_cf_handle_ = new ColumnFamilyHandleImpl(
versions_->GetColumnFamilySet()->GetDefault(), this, &mutex_);
default_cf_internal_stats_ = default_cf_handle_->cfd()->internal_stats();
single_column_family_mode_ =
versions_->GetColumnFamilySet()->NumberOfColumnFamilies() == 1;
std::unordered_set<ColumnFamilyData*> cfds_changed;
s = FindAndRecoverLogFiles(&cfds_changed, &job_context);
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
}
if (s.IsPathNotFound()) {
ROCKS_LOG_INFO(immutable_db_options_.info_log,
"Secondary tries to read WAL, but WAL file(s) have already "
"been purged by primary.");
s = Status::OK();
}
// TODO: update options_file_number_ needed?
job_context.Clean();
return s;
}
// find new WAL and apply them in order to the secondary instance
Status DBImplSecondary::FindAndRecoverLogFiles(
std::unordered_set<ColumnFamilyData*>* cfds_changed,
JobContext* job_context) {
assert(nullptr != cfds_changed);
assert(nullptr != job_context);
Status s;
std::vector<uint64_t> logs;
s = FindNewLogNumbers(&logs);
if (s.ok() && !logs.empty()) {
SequenceNumber next_sequence(kMaxSequenceNumber);
s = RecoverLogFiles(logs, &next_sequence, cfds_changed, job_context);
}
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
return s;
}
// List wal_dir and find all new WALs, return these log numbers
Status DBImplSecondary::FindNewLogNumbers(std::vector<uint64_t>* logs) {
assert(logs != nullptr);
std::vector<std::string> filenames;
Status s;
s = env_->GetChildren(immutable_db_options_.wal_dir, &filenames);
if (s.IsNotFound()) {
return Status::InvalidArgument("Failed to open wal_dir",
immutable_db_options_.wal_dir);
} else if (!s.ok()) {
return s;
}
// if log_readers_ is non-empty, it means we have applied all logs with log
// numbers smaller than the smallest log in log_readers_, so there is no
// need to pass these logs to RecoverLogFiles
uint64_t log_number_min = 0;
if (!log_readers_.empty()) {
log_number_min = log_readers_.begin()->first;
}
for (size_t i = 0; i < filenames.size(); i++) {
uint64_t number;
FileType type;
if (ParseFileName(filenames[i], &number, &type) && type == kLogFile &&
number >= log_number_min) {
logs->push_back(number);
}
}
// Recover logs in the order that they were generated
if (!logs->empty()) {
std::sort(logs->begin(), logs->end());
}
return s;
}
Status DBImplSecondary::MaybeInitLogReader(
uint64_t log_number, log::FragmentBufferedReader** log_reader) {
auto iter = log_readers_.find(log_number);
// make sure the log file is still present
if (iter == log_readers_.end() ||
iter->second->reader_->GetLogNumber() != log_number) {
// delete the obsolete log reader if log number mismatch
if (iter != log_readers_.end()) {
log_readers_.erase(iter);
}
// initialize log reader from log_number
// TODO: min_log_number_to_keep_2pc check needed?
// Open the log file
std::string fname = LogFileName(immutable_db_options_.wal_dir, log_number);
ROCKS_LOG_INFO(immutable_db_options_.info_log,
"Recovering log #%" PRIu64 " mode %d", log_number,
static_cast<int>(immutable_db_options_.wal_recovery_mode));
std::unique_ptr<SequentialFileReader> file_reader;
{
Introduce a new storage specific Env API (#5761) Summary: The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc. This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO. The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before. This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection. The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761 Differential Revision: D18868376 Pulled By: anand1976 fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
5 years ago
std::unique_ptr<FSSequentialFile> file;
Status status = fs_->NewSequentialFile(
fname, fs_->OptimizeForLogRead(file_options_), &file,
nullptr);
if (!status.ok()) {
*log_reader = nullptr;
return status;
}
file_reader.reset(new SequentialFileReader(
std::move(file), fname, immutable_db_options_.log_readahead_size));
}
// Create the log reader.
LogReaderContainer* log_reader_container = new LogReaderContainer(
env_, immutable_db_options_.info_log, std::move(fname),
std::move(file_reader), log_number);
log_readers_.insert(std::make_pair(
log_number, std::unique_ptr<LogReaderContainer>(log_reader_container)));
}
iter = log_readers_.find(log_number);
assert(iter != log_readers_.end());
*log_reader = iter->second->reader_;
return Status::OK();
}
// After manifest recovery, replay WALs and refresh log_readers_ if necessary
// REQUIRES: log_numbers are sorted in ascending order
Status DBImplSecondary::RecoverLogFiles(
const std::vector<uint64_t>& log_numbers, SequenceNumber* next_sequence,
std::unordered_set<ColumnFamilyData*>* cfds_changed,
JobContext* job_context) {
assert(nullptr != cfds_changed);
assert(nullptr != job_context);
mutex_.AssertHeld();
Status status;
for (auto log_number : log_numbers) {
log::FragmentBufferedReader* reader = nullptr;
status = MaybeInitLogReader(log_number, &reader);
if (!status.ok()) {
return status;
}
assert(reader != nullptr);
}
for (auto log_number : log_numbers) {
auto it = log_readers_.find(log_number);
assert(it != log_readers_.end());
log::FragmentBufferedReader* reader = it->second->reader_;
// Manually update the file number allocation counter in VersionSet.
versions_->MarkFileNumberUsed(log_number);
// Determine if we should tolerate incomplete records at the tail end of the
// Read all the records and add to a memtable
std::string scratch;
Slice record;
WriteBatch batch;
while (reader->ReadRecord(&record, &scratch,
immutable_db_options_.wal_recovery_mode) &&
status.ok()) {
if (record.size() < WriteBatchInternal::kHeader) {
reader->GetReporter()->Corruption(
record.size(), Status::Corruption("log record too small"));
continue;
}
WriteBatchInternal::SetContents(&batch, record);
SequenceNumber seq_of_batch = WriteBatchInternal::Sequence(&batch);
std::vector<uint32_t> column_family_ids;
status = CollectColumnFamilyIdsFromWriteBatch(batch, &column_family_ids);
if (status.ok()) {
for (const auto id : column_family_ids) {
ColumnFamilyData* cfd =
versions_->GetColumnFamilySet()->GetColumnFamily(id);
if (cfd == nullptr) {
continue;
}
if (cfds_changed->count(cfd) == 0) {
cfds_changed->insert(cfd);
}
Improve memtable earliest seqno assignment for secondary instance (#5413) Summary: In regular RocksDB instance, `MemTable::earliest_seqno_` is "db sequence number at the time of creation". However, we cannot use the db sequence number to set the value of `MemTable::earliest_seqno_` for secondary instance, i.e. `DBImplSecondary` due to the logic of MANIFEST and WAL replay. When replaying the log files of the primary, the secondary instance first replays MANIFEST and updates the db sequence number if necessary. Next, the secondary replays WAL files, creates new memtables if necessary and inserts key-value pairs into memtables. The following can occur when the db has two or more column families. Assume the db has column family "default" and "cf1". At a certain in time, both "default" and "cf1" have data in memtables. 1. Primary triggers a flush and flushes "cf1". "default" is **not** flushed. 2. Secondary replays the MANIFEST updates its db sequence number to the latest value learned from the MANIFEST. 3. Secondary starts to replay WAL that contains the writes to "default". It is possible that the write batches' sequence numbers are smaller than the db sequence number. In this case, these write batches will be skipped, and these updates will not be visible to reader until "default" is later flushed. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5413 Differential Revision: D15637407 Pulled By: riversand963 fbshipit-source-id: 3de3fe35cfc6f1b9f844f3f926f0df29717b6580
5 years ago
const std::vector<FileMetaData*>& l0_files =
cfd->current()->storage_info()->LevelFiles(0);
SequenceNumber seq =
l0_files.empty() ? 0 : l0_files.back()->fd.largest_seqno;
// If the write batch's sequence number is smaller than the last
// sequence number of the largest sequence persisted for this column
// family, then its data must reside in an SST that has already been
// added in the prior MANIFEST replay.
if (seq_of_batch <= seq) {
continue;
}
auto curr_log_num = port::kMaxUint64;
if (cfd_to_current_log_.count(cfd) > 0) {
curr_log_num = cfd_to_current_log_[cfd];
}
// If the active memtable contains records added by replaying an
// earlier WAL, then we need to seal the memtable, add it to the
// immutable memtable list and create a new active memtable.
if (!cfd->mem()->IsEmpty() && (curr_log_num == port::kMaxUint64 ||
curr_log_num != log_number)) {
const MutableCFOptions mutable_cf_options =
*cfd->GetLatestMutableCFOptions();
MemTable* new_mem =
Improve memtable earliest seqno assignment for secondary instance (#5413) Summary: In regular RocksDB instance, `MemTable::earliest_seqno_` is "db sequence number at the time of creation". However, we cannot use the db sequence number to set the value of `MemTable::earliest_seqno_` for secondary instance, i.e. `DBImplSecondary` due to the logic of MANIFEST and WAL replay. When replaying the log files of the primary, the secondary instance first replays MANIFEST and updates the db sequence number if necessary. Next, the secondary replays WAL files, creates new memtables if necessary and inserts key-value pairs into memtables. The following can occur when the db has two or more column families. Assume the db has column family "default" and "cf1". At a certain in time, both "default" and "cf1" have data in memtables. 1. Primary triggers a flush and flushes "cf1". "default" is **not** flushed. 2. Secondary replays the MANIFEST updates its db sequence number to the latest value learned from the MANIFEST. 3. Secondary starts to replay WAL that contains the writes to "default". It is possible that the write batches' sequence numbers are smaller than the db sequence number. In this case, these write batches will be skipped, and these updates will not be visible to reader until "default" is later flushed. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5413 Differential Revision: D15637407 Pulled By: riversand963 fbshipit-source-id: 3de3fe35cfc6f1b9f844f3f926f0df29717b6580
5 years ago
cfd->ConstructNewMemtable(mutable_cf_options, seq_of_batch);
cfd->mem()->SetNextLogNumber(log_number);
cfd->imm()->Add(cfd->mem(), &job_context->memtables_to_free);
new_mem->Ref();
cfd->SetMemtable(new_mem);
}
}
bool has_valid_writes = false;
status = WriteBatchInternal::InsertInto(
&batch, column_family_memtables_.get(),
Refactor trimming logic for immutable memtables (#5022) Summary: MyRocks currently sets `max_write_buffer_number_to_maintain` in order to maintain enough history for transaction conflict checking. The effectiveness of this approach depends on the size of memtables. When memtables are small, it may not keep enough history; when memtables are large, this may consume too much memory. We are proposing a new way to configure memtable list history: by limiting the memory usage of immutable memtables. The new option is `max_write_buffer_size_to_maintain` and it will take precedence over the old `max_write_buffer_number_to_maintain` if they are both set to non-zero values. The new option accounts for the total memory usage of flushed immutable memtables and mutable memtable. When the total usage exceeds the limit, RocksDB may start dropping immutable memtables (which is also called trimming history), starting from the oldest one. The semantics of the old option actually works both as an upper bound and lower bound. History trimming will start if number of immutable memtables exceeds the limit, but it will never go below (limit-1) due to history trimming. In order the mimic the behavior with the new option, history trimming will stop if dropping the next immutable memtable causes the total memory usage go below the size limit. For example, assuming the size limit is set to 64MB, and there are 3 immutable memtables with sizes of 20, 30, 30. Although the total memory usage is 80MB > 64MB, dropping the oldest memtable will reduce the memory usage to 60MB < 64MB, so in this case no memtable will be dropped. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5022 Differential Revision: D14394062 Pulled By: miasantreble fbshipit-source-id: 60457a509c6af89d0993f988c9b5c2aa9e45f5c5
5 years ago
nullptr /* flush_scheduler */, nullptr /* trim_history_scheduler*/,
true, log_number, this, false /* concurrent_memtable_writes */,
next_sequence, &has_valid_writes, seq_per_batch_, batch_per_txn_);
}
// If column family was not found, it might mean that the WAL write
// batch references to the column family that was dropped after the
// insert. We don't want to fail the whole write batch in that case --
// we just ignore the update.
// That's why we set ignore missing column families to true
// passing null flush_scheduler will disable memtable flushing which is
// needed for secondary instances
if (status.ok()) {
for (const auto id : column_family_ids) {
ColumnFamilyData* cfd =
versions_->GetColumnFamilySet()->GetColumnFamily(id);
if (cfd == nullptr) {
continue;
}
std::unordered_map<ColumnFamilyData*, uint64_t>::iterator iter =
cfd_to_current_log_.find(cfd);
if (iter == cfd_to_current_log_.end()) {
cfd_to_current_log_.insert({cfd, log_number});
} else if (log_number > iter->second) {
iter->second = log_number;
}
}
auto last_sequence = *next_sequence - 1;
if ((*next_sequence != kMaxSequenceNumber) &&
(versions_->LastSequence() <= last_sequence)) {
versions_->SetLastAllocatedSequence(last_sequence);
versions_->SetLastPublishedSequence(last_sequence);
versions_->SetLastSequence(last_sequence);
}
} else {
// We are treating this as a failure while reading since we read valid
// blocks that do not form coherent data
reader->GetReporter()->Corruption(record.size(), status);
}
}
if (!status.ok()) {
return status;
}
}
// remove logreaders from map after successfully recovering the WAL
if (log_readers_.size() > 1) {
auto erase_iter = log_readers_.begin();
std::advance(erase_iter, log_readers_.size() - 1);
log_readers_.erase(log_readers_.begin(), erase_iter);
}
return status;
}
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
// Implementation of the DB interface
Status DBImplSecondary::Get(const ReadOptions& read_options,
ColumnFamilyHandle* column_family, const Slice& key,
PinnableSlice* value) {
return GetImpl(read_options, column_family, key, value);
}
Status DBImplSecondary::GetImpl(const ReadOptions& read_options,
ColumnFamilyHandle* column_family,
const Slice& key, PinnableSlice* pinnable_val) {
assert(pinnable_val != nullptr);
PERF_CPU_TIMER_GUARD(get_cpu_nanos, env_);
StopWatch sw(env_, stats_, DB_GET);
PERF_TIMER_GUARD(get_snapshot_time);
auto cfh = static_cast<ColumnFamilyHandleImpl*>(column_family);
ColumnFamilyData* cfd = cfh->cfd();
if (tracer_) {
InstrumentedMutexLock lock(&trace_mutex_);
if (tracer_) {
tracer_->Get(column_family, key);
}
}
// Acquire SuperVersion
SuperVersion* super_version = GetAndRefSuperVersion(cfd);
SequenceNumber snapshot = versions_->LastSequence();
MergeContext merge_context;
SequenceNumber max_covering_tombstone_seq = 0;
Status s;
LookupKey lkey(key, snapshot);
PERF_TIMER_STOP(get_snapshot_time);
bool done = false;
return timestamp from get (#6409) Summary: Added new Get() methods that return timestamp. Dummy implementation is given so that classes derived from DB don't need to be touched to provide their implementation. MultiGet is not included. ReadRandom perf test (10 minutes) on the same development machine ram drive with the same DB data shows no regression (within marge of error). The test is adapted from https://github.com/facebook/rocksdb/wiki/RocksDB-In-Memory-Workload-Performance-Benchmarks. base line (commit 72ee067b9): 101.712 micros/op 314602 ops/sec; 36.0 MB/s (5658999 of 5658999 found) This PR: 100.288 micros/op 319071 ops/sec; 36.5 MB/s (5674999 of 5674999 found) ./db_bench --db=r:\rocksdb.github --num_levels=6 --key_size=20 --prefix_size=20 --keys_per_prefix=0 --value_size=100 --cache_size=2147483648 --cache_numshardbits=6 --compression_type=none --compression_ratio=1 --min_level_to_compress=-1 --disable_seek_compaction=1 --hard_rate_limit=2 --write_buffer_size=134217728 --max_write_buffer_number=2 --level0_file_num_compaction_trigger=8 --target_file_size_base=134217728 --max_bytes_for_level_base=1073741824 --disable_wal=0 --wal_dir=r:\rocksdb.github\WAL_LOG --sync=0 --verify_checksum=1 --delete_obsolete_files_period_micros=314572800 --max_background_compactions=4 --max_background_flushes=0 --level0_slowdown_writes_trigger=16 --level0_stop_writes_trigger=24 --statistics=0 --stats_per_interval=0 --stats_interval=1048576 --histogram=0 --use_plain_table=1 --open_files=-1 --mmap_read=1 --mmap_write=0 --memtablerep=prefix_hash --bloom_bits=10 --bloom_locality=1 --duration=600 --benchmarks=readrandom --use_existing_db=1 --num=25000000 --threads=32 Pull Request resolved: https://github.com/facebook/rocksdb/pull/6409 Differential Revision: D20200086 Pulled By: riversand963 fbshipit-source-id: 490edd74d924f62bd8ae9c29c2a6bbbb8410ca50
5 years ago
if (super_version->mem->Get(lkey, pinnable_val->GetSelf(),
/*timestamp=*/nullptr, &s, &merge_context,
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
&max_covering_tombstone_seq, read_options)) {
done = true;
pinnable_val->PinSelf();
RecordTick(stats_, MEMTABLE_HIT);
} else if ((s.ok() || s.IsMergeInProgress()) &&
super_version->imm->Get(
return timestamp from get (#6409) Summary: Added new Get() methods that return timestamp. Dummy implementation is given so that classes derived from DB don't need to be touched to provide their implementation. MultiGet is not included. ReadRandom perf test (10 minutes) on the same development machine ram drive with the same DB data shows no regression (within marge of error). The test is adapted from https://github.com/facebook/rocksdb/wiki/RocksDB-In-Memory-Workload-Performance-Benchmarks. base line (commit 72ee067b9): 101.712 micros/op 314602 ops/sec; 36.0 MB/s (5658999 of 5658999 found) This PR: 100.288 micros/op 319071 ops/sec; 36.5 MB/s (5674999 of 5674999 found) ./db_bench --db=r:\rocksdb.github --num_levels=6 --key_size=20 --prefix_size=20 --keys_per_prefix=0 --value_size=100 --cache_size=2147483648 --cache_numshardbits=6 --compression_type=none --compression_ratio=1 --min_level_to_compress=-1 --disable_seek_compaction=1 --hard_rate_limit=2 --write_buffer_size=134217728 --max_write_buffer_number=2 --level0_file_num_compaction_trigger=8 --target_file_size_base=134217728 --max_bytes_for_level_base=1073741824 --disable_wal=0 --wal_dir=r:\rocksdb.github\WAL_LOG --sync=0 --verify_checksum=1 --delete_obsolete_files_period_micros=314572800 --max_background_compactions=4 --max_background_flushes=0 --level0_slowdown_writes_trigger=16 --level0_stop_writes_trigger=24 --statistics=0 --stats_per_interval=0 --stats_interval=1048576 --histogram=0 --use_plain_table=1 --open_files=-1 --mmap_read=1 --mmap_write=0 --memtablerep=prefix_hash --bloom_bits=10 --bloom_locality=1 --duration=600 --benchmarks=readrandom --use_existing_db=1 --num=25000000 --threads=32 Pull Request resolved: https://github.com/facebook/rocksdb/pull/6409 Differential Revision: D20200086 Pulled By: riversand963 fbshipit-source-id: 490edd74d924f62bd8ae9c29c2a6bbbb8410ca50
5 years ago
lkey, pinnable_val->GetSelf(), /*timestamp=*/nullptr, &s,
&merge_context, &max_covering_tombstone_seq, read_options)) {
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
done = true;
pinnable_val->PinSelf();
RecordTick(stats_, MEMTABLE_HIT);
}
if (!done && !s.ok() && !s.IsMergeInProgress()) {
ReturnAndCleanupSuperVersion(cfd, super_version);
return s;
}
if (!done) {
PERF_TIMER_GUARD(get_from_output_files_time);
return timestamp from get (#6409) Summary: Added new Get() methods that return timestamp. Dummy implementation is given so that classes derived from DB don't need to be touched to provide their implementation. MultiGet is not included. ReadRandom perf test (10 minutes) on the same development machine ram drive with the same DB data shows no regression (within marge of error). The test is adapted from https://github.com/facebook/rocksdb/wiki/RocksDB-In-Memory-Workload-Performance-Benchmarks. base line (commit 72ee067b9): 101.712 micros/op 314602 ops/sec; 36.0 MB/s (5658999 of 5658999 found) This PR: 100.288 micros/op 319071 ops/sec; 36.5 MB/s (5674999 of 5674999 found) ./db_bench --db=r:\rocksdb.github --num_levels=6 --key_size=20 --prefix_size=20 --keys_per_prefix=0 --value_size=100 --cache_size=2147483648 --cache_numshardbits=6 --compression_type=none --compression_ratio=1 --min_level_to_compress=-1 --disable_seek_compaction=1 --hard_rate_limit=2 --write_buffer_size=134217728 --max_write_buffer_number=2 --level0_file_num_compaction_trigger=8 --target_file_size_base=134217728 --max_bytes_for_level_base=1073741824 --disable_wal=0 --wal_dir=r:\rocksdb.github\WAL_LOG --sync=0 --verify_checksum=1 --delete_obsolete_files_period_micros=314572800 --max_background_compactions=4 --max_background_flushes=0 --level0_slowdown_writes_trigger=16 --level0_stop_writes_trigger=24 --statistics=0 --stats_per_interval=0 --stats_interval=1048576 --histogram=0 --use_plain_table=1 --open_files=-1 --mmap_read=1 --mmap_write=0 --memtablerep=prefix_hash --bloom_bits=10 --bloom_locality=1 --duration=600 --benchmarks=readrandom --use_existing_db=1 --num=25000000 --threads=32 Pull Request resolved: https://github.com/facebook/rocksdb/pull/6409 Differential Revision: D20200086 Pulled By: riversand963 fbshipit-source-id: 490edd74d924f62bd8ae9c29c2a6bbbb8410ca50
5 years ago
super_version->current->Get(read_options, lkey, pinnable_val,
/*timestamp=*/nullptr, &s, &merge_context,
&max_covering_tombstone_seq);
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
RecordTick(stats_, MEMTABLE_MISS);
}
{
PERF_TIMER_GUARD(get_post_process_time);
ReturnAndCleanupSuperVersion(cfd, super_version);
RecordTick(stats_, NUMBER_KEYS_READ);
size_t size = pinnable_val->size();
RecordTick(stats_, BYTES_READ, size);
RecordTimeToHistogram(stats_, BYTES_PER_READ, size);
PERF_COUNTER_ADD(get_read_bytes, size);
}
return s;
}
Iterator* DBImplSecondary::NewIterator(const ReadOptions& read_options,
ColumnFamilyHandle* column_family) {
if (read_options.managed) {
return NewErrorIterator(
Status::NotSupported("Managed iterator is not supported anymore."));
}
if (read_options.read_tier == kPersistedTier) {
return NewErrorIterator(Status::NotSupported(
"ReadTier::kPersistedData is not yet supported in iterators."));
}
Iterator* result = nullptr;
auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
auto cfd = cfh->cfd();
ReadCallback* read_callback = nullptr; // No read callback provided.
if (read_options.tailing) {
return NewErrorIterator(Status::NotSupported(
"tailing iterator not supported in secondary mode"));
} else if (read_options.snapshot != nullptr) {
// TODO (yanqin) support snapshot.
return NewErrorIterator(
Status::NotSupported("snapshot not supported in secondary mode"));
} else {
auto snapshot = versions_->LastSequence();
result = NewIteratorImpl(read_options, cfd, snapshot, read_callback);
}
return result;
}
ArenaWrappedDBIter* DBImplSecondary::NewIteratorImpl(
const ReadOptions& read_options, ColumnFamilyData* cfd,
SequenceNumber snapshot, ReadCallback* read_callback) {
assert(nullptr != cfd);
SuperVersion* super_version = cfd->GetReferencedSuperVersion(this);
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
auto db_iter = NewArenaWrappedDbIterator(
env_, read_options, *cfd->ioptions(), super_version->mutable_cf_options,
snapshot,
super_version->mutable_cf_options.max_sequential_skip_in_iterations,
super_version->version_number, read_callback);
auto internal_iter =
NewInternalIterator(read_options, cfd, super_version, db_iter->GetArena(),
Properly report IO errors when IndexType::kBinarySearchWithFirstKey is used (#6621) Summary: Context: Index type `kBinarySearchWithFirstKey` added the ability for sst file iterator to sometimes report a key from index without reading the corresponding data block. This is useful when sst blocks are cut at some meaningful boundaries (e.g. one block per key prefix), and many seeks land between blocks (e.g. for each prefix, the ranges of keys in different sst files are nearly disjoint, so a typical seek needs to read a data block from only one file even if all files have the prefix). But this added a new error condition, which rocksdb code was really not equipped to deal with: `InternalIterator::value()` may fail with an IO error or Status::Incomplete, but it's just a method returning a Slice, with no way to report error instead. Before this PR, this type of error wasn't handled at all (an empty slice was returned), and kBinarySearchWithFirstKey implementation was considered a prototype. Now that we (LogDevice) have experimented with kBinarySearchWithFirstKey for a while and confirmed that it's really useful, this PR is adding the missing error handling. It's a pretty inconvenient situation implementation-wise. The error needs to be reported from InternalIterator when trying to access value. But there are ~700 call sites of `InternalIterator::value()`, most of which either can't hit the error condition (because the iterator is reading from memtable or from index or something) or wouldn't benefit from the deferred loading of the value (e.g. compaction iterator that reads all values anyway). Adding error handling to all these call sites would needlessly bloat the code. So instead I made the deferred value loading optional: only the call sites that may use deferred loading have to call the new method `PrepareValue()` before calling `value()`. The feature is enabled with a new bool argument `allow_unprepared_value` to a bunch of methods that create iterators (it wouldn't make sense to put it in ReadOptions because it's completely internal to iterators, with virtually no user-visible effect). Lmk if you have better ideas. Note that the deferred value loading only happens for *internal* iterators. The user-visible iterator (DBIter) always prepares the value before returning from Seek/Next/etc. We could go further and add an API to defer that value loading too, but that's most likely not useful for LogDevice, so it doesn't seem worth the complexity for now. Pull Request resolved: https://github.com/facebook/rocksdb/pull/6621 Test Plan: make -j5 check . Will also deploy to some logdevice test clusters and look at stats. Reviewed By: siying Differential Revision: D20786930 Pulled By: al13n321 fbshipit-source-id: 6da77d918bad3780522e918f17f4d5513d3e99ee
5 years ago
db_iter->GetRangeDelAggregator(), snapshot,
/* allow_unprepared_value */ true);
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
db_iter->SetIterUnderDBIter(internal_iter);
return db_iter;
}
Status DBImplSecondary::NewIterators(
const ReadOptions& read_options,
const std::vector<ColumnFamilyHandle*>& column_families,
std::vector<Iterator*>* iterators) {
if (read_options.managed) {
return Status::NotSupported("Managed iterator is not supported anymore.");
}
if (read_options.read_tier == kPersistedTier) {
return Status::NotSupported(
"ReadTier::kPersistedData is not yet supported in iterators.");
}
ReadCallback* read_callback = nullptr; // No read callback provided.
if (iterators == nullptr) {
return Status::InvalidArgument("iterators not allowed to be nullptr");
}
iterators->clear();
iterators->reserve(column_families.size());
if (read_options.tailing) {
return Status::NotSupported(
"tailing iterator not supported in secondary mode");
} else if (read_options.snapshot != nullptr) {
// TODO (yanqin) support snapshot.
return Status::NotSupported("snapshot not supported in secondary mode");
} else {
SequenceNumber read_seq = versions_->LastSequence();
for (auto cfh : column_families) {
ColumnFamilyData* cfd = static_cast<ColumnFamilyHandleImpl*>(cfh)->cfd();
iterators->push_back(
NewIteratorImpl(read_options, cfd, read_seq, read_callback));
}
}
return Status::OK();
}
Status DBImplSecondary::CheckConsistency() {
mutex_.AssertHeld();
Status s = DBImpl::CheckConsistency();
// If DBImpl::CheckConsistency() which is stricter returns success, then we
// do not need to give a second chance.
if (s.ok()) {
return s;
}
// It's possible that DBImpl::CheckConssitency() can fail because the primary
// may have removed certain files, causing the GetFileSize(name) call to
// fail and returning a PathNotFound. In this case, we take a best-effort
// approach and just proceed.
TEST_SYNC_POINT_CALLBACK(
"DBImplSecondary::CheckConsistency:AfterFirstAttempt", &s);
if (immutable_db_options_.skip_checking_sst_file_sizes_on_db_open) {
return Status::OK();
}
std::vector<LiveFileMetaData> metadata;
versions_->GetLiveFilesMetaData(&metadata);
std::string corruption_messages;
for (const auto& md : metadata) {
// md.name has a leading "/".
std::string file_path = md.db_path + md.name;
uint64_t fsize = 0;
s = env_->GetFileSize(file_path, &fsize);
if (!s.ok() &&
(env_->GetFileSize(Rocks2LevelTableFileName(file_path), &fsize).ok() ||
s.IsPathNotFound())) {
s = Status::OK();
}
if (!s.ok()) {
corruption_messages +=
"Can't access " + md.name + ": " + s.ToString() + "\n";
}
}
return corruption_messages.empty() ? Status::OK()
: Status::Corruption(corruption_messages);
}
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
Status DBImplSecondary::TryCatchUpWithPrimary() {
assert(versions_.get() != nullptr);
assert(manifest_reader_.get() != nullptr);
Status s;
// read the manifest and apply new changes to the secondary instance
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
std::unordered_set<ColumnFamilyData*> cfds_changed;
JobContext job_context(0, true /*create_superversion*/);
{
InstrumentedMutexLock lock_guard(&mutex_);
s = static_cast_with_check<ReactiveVersionSet>(versions_.get())
->ReadAndApply(&mutex_, &manifest_reader_, &cfds_changed);
ROCKS_LOG_INFO(immutable_db_options_.info_log, "Last sequence is %" PRIu64,
static_cast<uint64_t>(versions_->LastSequence()));
for (ColumnFamilyData* cfd : cfds_changed) {
if (cfd->IsDropped()) {
ROCKS_LOG_DEBUG(immutable_db_options_.info_log, "[%s] is dropped\n",
cfd->GetName().c_str());
continue;
}
VersionStorageInfo::LevelSummaryStorage tmp;
ROCKS_LOG_DEBUG(immutable_db_options_.info_log,
"[%s] Level summary: %s\n", cfd->GetName().c_str(),
cfd->current()->storage_info()->LevelSummary(&tmp));
Improve memtable earliest seqno assignment for secondary instance (#5413) Summary: In regular RocksDB instance, `MemTable::earliest_seqno_` is "db sequence number at the time of creation". However, we cannot use the db sequence number to set the value of `MemTable::earliest_seqno_` for secondary instance, i.e. `DBImplSecondary` due to the logic of MANIFEST and WAL replay. When replaying the log files of the primary, the secondary instance first replays MANIFEST and updates the db sequence number if necessary. Next, the secondary replays WAL files, creates new memtables if necessary and inserts key-value pairs into memtables. The following can occur when the db has two or more column families. Assume the db has column family "default" and "cf1". At a certain in time, both "default" and "cf1" have data in memtables. 1. Primary triggers a flush and flushes "cf1". "default" is **not** flushed. 2. Secondary replays the MANIFEST updates its db sequence number to the latest value learned from the MANIFEST. 3. Secondary starts to replay WAL that contains the writes to "default". It is possible that the write batches' sequence numbers are smaller than the db sequence number. In this case, these write batches will be skipped, and these updates will not be visible to reader until "default" is later flushed. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5413 Differential Revision: D15637407 Pulled By: riversand963 fbshipit-source-id: 3de3fe35cfc6f1b9f844f3f926f0df29717b6580
5 years ago
}
// list wal_dir to discover new WALs and apply new changes to the secondary
// instance
if (s.ok()) {
s = FindAndRecoverLogFiles(&cfds_changed, &job_context);
}
if (s.IsPathNotFound()) {
ROCKS_LOG_INFO(
immutable_db_options_.info_log,
"Secondary tries to read WAL, but WAL file(s) have already "
"been purged by primary.");
s = Status::OK();
}
if (s.ok()) {
for (auto cfd : cfds_changed) {
cfd->imm()->RemoveOldMemTables(cfd->GetLogNumber(),
&job_context.memtables_to_free);
auto& sv_context = job_context.superversion_contexts.back();
cfd->InstallSuperVersion(&sv_context, &mutex_);
sv_context.NewSuperVersion();
}
}
}
job_context.Clean();
// Cleanup unused, obsolete files.
JobContext purge_files_job_context(0);
{
InstrumentedMutexLock lock_guard(&mutex_);
// Currently, secondary instance does not own the database files, thus it
// is unnecessary for the secondary to force full scan.
FindObsoleteFiles(&purge_files_job_context, /*force=*/false);
}
if (purge_files_job_context.HaveSomethingToDelete()) {
PurgeObsoleteFiles(purge_files_job_context);
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
}
purge_files_job_context.Clean();
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
return s;
}
Status DB::OpenAsSecondary(const Options& options, const std::string& dbname,
const std::string& secondary_path, DB** dbptr) {
*dbptr = nullptr;
DBOptions db_options(options);
ColumnFamilyOptions cf_options(options);
std::vector<ColumnFamilyDescriptor> column_families;
column_families.emplace_back(kDefaultColumnFamilyName, cf_options);
std::vector<ColumnFamilyHandle*> handles;
Status s = DB::OpenAsSecondary(db_options, dbname, secondary_path,
column_families, &handles, dbptr);
if (s.ok()) {
assert(handles.size() == 1);
delete handles[0];
}
return s;
}
Status DB::OpenAsSecondary(
const DBOptions& db_options, const std::string& dbname,
const std::string& secondary_path,
const std::vector<ColumnFamilyDescriptor>& column_families,
std::vector<ColumnFamilyHandle*>* handles, DB** dbptr) {
*dbptr = nullptr;
if (db_options.max_open_files != -1) {
// TODO (yanqin) maybe support max_open_files != -1 by creating hard links
// on SST files so that db secondary can still have access to old SSTs
// while primary instance may delete original.
return Status::InvalidArgument("require max_open_files to be -1");
}
DBOptions tmp_opts(db_options);
Status s;
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
if (nullptr == tmp_opts.info_log) {
s = CreateLoggerFromOptions(secondary_path, tmp_opts, &tmp_opts.info_log);
if (!s.ok()) {
tmp_opts.info_log = nullptr;
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
}
}
handles->clear();
DBImplSecondary* impl = new DBImplSecondary(tmp_opts, dbname);
impl->versions_.reset(new ReactiveVersionSet(
Introduce a new storage specific Env API (#5761) Summary: The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc. This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO. The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before. This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection. The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761 Differential Revision: D18868376 Pulled By: anand1976 fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
5 years ago
dbname, &impl->immutable_db_options_, impl->file_options_,
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
impl->table_cache_.get(), impl->write_buffer_manager_,
&impl->write_controller_));
impl->column_family_memtables_.reset(
new ColumnFamilyMemTablesImpl(impl->versions_->GetColumnFamilySet()));
impl->wal_in_db_path_ = IsWalDirSameAsDBPath(&impl->immutable_db_options_);
setup wal_in_db_path_ for secondary instance (#5545) Summary: PR https://github.com/facebook/rocksdb/pull/5520 adds DBImpl:: wal_in_db_path_ and initializes it in DBImpl::Open, this PR fixes the valgrind error for secondary instance: ``` ==236417== Conditional jump or move depends on uninitialised value(s) ==236417== at 0x62242A: rocksdb::DeleteDBFile(rocksdb::ImmutableDBOptions const*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, bool, bool) (file_util.cc:96) ==236417== by 0x512432: rocksdb::DBImpl::DeleteObsoleteFileImpl(int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, rocksdb::FileType, unsigned long) (db_impl_files.cc:261) ==236417== by 0x515A7A: rocksdb::DBImpl::PurgeObsoleteFiles(rocksdb::JobContext&, bool) (db_impl_files.cc:492) ==236417== by 0x499153: rocksdb::ColumnFamilyHandleImpl::~ColumnFamilyHandleImpl() (column_family.cc:75) ==236417== by 0x499880: rocksdb::ColumnFamilyHandleImpl::~ColumnFamilyHandleImpl() (column_family.cc:84) ==236417== by 0x4C9AF9: rocksdb::DB::DestroyColumnFamilyHandle(rocksdb::ColumnFamilyHandle*) (db_impl.cc:3105) ==236417== by 0x44E853: CloseSecondary (db_secondary_test.cc:53) ==236417== by 0x44E853: rocksdb::DBSecondaryTest::~DBSecondaryTest() (db_secondary_test.cc:31) ==236417== by 0x44EC77: ~DBSecondaryTest_PrimaryDropColumnFamily_Test (db_secondary_test.cc:443) ==236417== by 0x44EC77: rocksdb::DBSecondaryTest_PrimaryDropColumnFamily_Test::~DBSecondaryTest_PrimaryDropColumnFamily_Test() (db_secondary_test.cc:443) ==236417== by 0x83D1D7: HandleSehExceptionsInMethodIfSupported<testing::Test, void> (gtest-all.cc:3824) ==236417== by 0x83D1D7: void testing::internal::HandleExceptionsInMethodIfSupported<testing::Test, void>(testing::Test*, void (testing::Test::*)(), char const*) (gtest-all.cc:3860) ==236417== by 0x8346DB: testing::TestInfo::Run() [clone .part.486] (gtest-all.cc:4078) ==236417== by 0x8348D4: Run (gtest-all.cc:4047) ==236417== by 0x8348D4: testing::TestCase::Run() [clone .part.487] (gtest-all.cc:4190) ==236417== by 0x834D14: Run (gtest-all.cc:6100) ==236417== by 0x834D14: testing::internal::UnitTestImpl::RunAllTests() (gtest-all.cc:6062) ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/5545 Differential Revision: D16146224 Pulled By: miasantreble fbshipit-source-id: 184c90e451352951da4e955f054d4b1a1f29ea29
5 years ago
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
impl->mutex_.Lock();
s = impl->Recover(column_families, true, false, false);
Support for single-primary, multi-secondary instances (#4899) Summary: This PR allows RocksDB to run in single-primary, multi-secondary process mode. The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary. Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`. This PR has several components: 1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary. 2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue. 3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`. 3.1 Tail the primary's MANIFEST during recovery. 3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`. 3.3 Tailing WAL will be in a future PR. 4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899 Differential Revision: D14510945 Pulled By: riversand963 fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
6 years ago
if (s.ok()) {
for (auto cf : column_families) {
auto cfd =
impl->versions_->GetColumnFamilySet()->GetColumnFamily(cf.name);
if (nullptr == cfd) {
s = Status::InvalidArgument("Column family not found: ", cf.name);
break;
}
handles->push_back(new ColumnFamilyHandleImpl(cfd, impl, &impl->mutex_));
}
}
SuperVersionContext sv_context(true /* create_superversion */);
if (s.ok()) {
for (auto cfd : *impl->versions_->GetColumnFamilySet()) {
sv_context.NewSuperVersion();
cfd->InstallSuperVersion(&sv_context, &impl->mutex_);
}
}
impl->mutex_.Unlock();
sv_context.Clean();
if (s.ok()) {
*dbptr = impl;
for (auto h : *handles) {
impl->NewThreadStatusCfInfo(
reinterpret_cast<ColumnFamilyHandleImpl*>(h)->cfd());
}
} else {
for (auto h : *handles) {
delete h;
}
handles->clear();
delete impl;
}
return s;
}
#else // !ROCKSDB_LITE
Status DB::OpenAsSecondary(const Options& /*options*/,
const std::string& /*name*/,
const std::string& /*secondary_path*/,
DB** /*dbptr*/) {
return Status::NotSupported("Not supported in ROCKSDB_LITE.");
}
Status DB::OpenAsSecondary(
const DBOptions& /*db_options*/, const std::string& /*dbname*/,
const std::string& /*secondary_path*/,
const std::vector<ColumnFamilyDescriptor>& /*column_families*/,
std::vector<ColumnFamilyHandle*>* /*handles*/, DB** /*dbptr*/) {
return Status::NotSupported("Not supported in ROCKSDB_LITE.");
}
#endif // !ROCKSDB_LITE
} // namespace ROCKSDB_NAMESPACE