2016-02-09 23:12:00 +00:00
|
|
|
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
|
2017-07-15 23:03:42 +00:00
|
|
|
// 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).
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
|
|
|
#pragma once
|
|
|
|
|
|
|
|
#ifndef ROCKSDB_LITE
|
|
|
|
|
2016-12-06 01:18:14 +00:00
|
|
|
#include <algorithm>
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
#include <atomic>
|
2016-09-28 00:43:06 +00:00
|
|
|
#include <mutex>
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
#include <stack>
|
|
|
|
#include <string>
|
|
|
|
#include <unordered_map>
|
|
|
|
#include <vector>
|
|
|
|
|
|
|
|
#include "db/write_callback.h"
|
|
|
|
#include "rocksdb/db.h"
|
|
|
|
#include "rocksdb/slice.h"
|
2015-07-31 22:18:27 +00:00
|
|
|
#include "rocksdb/snapshot.h"
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
#include "rocksdb/status.h"
|
|
|
|
#include "rocksdb/types.h"
|
|
|
|
#include "rocksdb/utilities/transaction.h"
|
|
|
|
#include "rocksdb/utilities/transaction_db.h"
|
|
|
|
#include "rocksdb/utilities/write_batch_with_index.h"
|
2016-12-06 01:18:14 +00:00
|
|
|
#include "util/autovector.h"
|
2015-08-21 22:47:21 +00:00
|
|
|
#include "utilities/transactions/transaction_base.h"
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
#include "utilities/transactions/transaction_util.h"
|
|
|
|
|
2020-02-20 20:07:53 +00:00
|
|
|
namespace ROCKSDB_NAMESPACE {
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-08-06 00:17:48 +00:00
|
|
|
class PessimisticTransactionDB;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-08-03 15:46:47 +00:00
|
|
|
// A transaction under pessimistic concurrency control. This class implements
|
|
|
|
// the locking API and interfaces with the lock manager as well as the
|
|
|
|
// pessimistic transactional db.
|
2017-08-07 23:07:40 +00:00
|
|
|
class PessimisticTransaction : public TransactionBaseImpl {
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
public:
|
2017-08-07 23:07:40 +00:00
|
|
|
PessimisticTransaction(TransactionDB* db, const WriteOptions& write_options,
|
2019-01-16 02:07:50 +00:00
|
|
|
const TransactionOptions& txn_options,
|
|
|
|
const bool init = true);
|
2019-09-12 01:07:12 +00:00
|
|
|
// No copying allowed
|
|
|
|
PessimisticTransaction(const PessimisticTransaction&) = delete;
|
|
|
|
void operator=(const PessimisticTransaction&) = delete;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-08-07 23:07:40 +00:00
|
|
|
virtual ~PessimisticTransaction();
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2016-03-03 23:36:26 +00:00
|
|
|
void Reinitialize(TransactionDB* txn_db, const WriteOptions& write_options,
|
2016-02-03 03:19:17 +00:00
|
|
|
const TransactionOptions& txn_options);
|
|
|
|
|
2017-08-07 23:07:40 +00:00
|
|
|
Status Prepare() override;
|
2016-04-18 18:15:50 +00:00
|
|
|
|
2017-08-07 23:07:40 +00:00
|
|
|
Status Commit() override;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-09-08 22:53:51 +00:00
|
|
|
// It is basically Commit without going through Prepare phase. The write batch
|
|
|
|
// is also directly provided instead of expecting txn to gradually batch the
|
|
|
|
// transactions writes to an internal write batch.
|
2017-09-13 18:56:27 +00:00
|
|
|
Status CommitBatch(WriteBatch* batch);
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-10-03 02:46:42 +00:00
|
|
|
Status Rollback() override;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2015-09-12 01:10:50 +00:00
|
|
|
Status RollbackToSavePoint() override;
|
2015-08-25 02:13:18 +00:00
|
|
|
|
2016-04-18 18:15:50 +00:00
|
|
|
Status SetName(const TransactionName& name) override;
|
|
|
|
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
// Generate a new unique transaction identifier
|
|
|
|
static TransactionID GenTxnID();
|
|
|
|
|
2016-09-28 00:43:06 +00:00
|
|
|
TransactionID GetID() const override { return txn_id_; }
|
|
|
|
|
2016-12-06 01:18:14 +00:00
|
|
|
std::vector<TransactionID> GetWaitingTxns(uint32_t* column_family_id,
|
|
|
|
std::string* key) const override {
|
2016-09-28 00:43:06 +00:00
|
|
|
std::lock_guard<std::mutex> lock(wait_mutex_);
|
2016-12-06 01:18:14 +00:00
|
|
|
std::vector<TransactionID> ids(waiting_txn_ids_.size());
|
|
|
|
if (key) *key = waiting_key_ ? *waiting_key_ : "";
|
2016-09-28 00:43:06 +00:00
|
|
|
if (column_family_id) *column_family_id = waiting_cf_id_;
|
2016-12-06 01:18:14 +00:00
|
|
|
std::copy(waiting_txn_ids_.begin(), waiting_txn_ids_.end(), ids.begin());
|
|
|
|
return ids;
|
2016-09-28 00:43:06 +00:00
|
|
|
}
|
|
|
|
|
2016-12-06 01:18:14 +00:00
|
|
|
void SetWaitingTxn(autovector<TransactionID> ids, uint32_t column_family_id,
|
2016-09-28 00:43:06 +00:00
|
|
|
const std::string* key) {
|
|
|
|
std::lock_guard<std::mutex> lock(wait_mutex_);
|
2016-12-06 01:18:14 +00:00
|
|
|
waiting_txn_ids_ = ids;
|
2016-09-28 00:43:06 +00:00
|
|
|
waiting_cf_id_ = column_family_id;
|
|
|
|
waiting_key_ = key;
|
|
|
|
}
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2016-12-06 01:18:14 +00:00
|
|
|
void ClearWaitingTxn() {
|
|
|
|
std::lock_guard<std::mutex> lock(wait_mutex_);
|
|
|
|
waiting_txn_ids_.clear();
|
|
|
|
waiting_cf_id_ = 0;
|
|
|
|
waiting_key_ = nullptr;
|
|
|
|
}
|
|
|
|
|
2015-09-08 19:36:48 +00:00
|
|
|
// Returns the time (in microseconds according to Env->GetMicros())
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
// that this transaction will be expired. Returns 0 if this transaction does
|
|
|
|
// not expire.
|
|
|
|
uint64_t GetExpirationTime() const { return expiration_time_; }
|
|
|
|
|
|
|
|
// returns true if this transaction has an expiration_time and has expired.
|
|
|
|
bool IsExpired() const;
|
|
|
|
|
2015-09-08 19:36:48 +00:00
|
|
|
// Returns the number of microseconds a transaction can wait on acquiring a
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
// lock or -1 if there is no timeout.
|
|
|
|
int64_t GetLockTimeout() const { return lock_timeout_; }
|
2015-09-08 19:36:48 +00:00
|
|
|
void SetLockTimeout(int64_t timeout) override {
|
|
|
|
lock_timeout_ = timeout * 1000;
|
|
|
|
}
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2016-02-02 01:07:05 +00:00
|
|
|
// Returns true if locks were stolen successfully, false otherwise.
|
|
|
|
bool TryStealingLocks();
|
|
|
|
|
2016-10-21 18:39:15 +00:00
|
|
|
bool IsDeadlockDetect() const override { return deadlock_detect_; }
|
2016-09-28 08:23:33 +00:00
|
|
|
|
|
|
|
int64_t GetDeadlockDetectDepth() const { return deadlock_detect_depth_; }
|
|
|
|
|
2015-08-21 22:47:21 +00:00
|
|
|
protected:
|
2017-11-02 00:23:52 +00:00
|
|
|
// Refer to
|
|
|
|
// TransactionOptions::use_only_the_last_commit_time_batch_for_recovery
|
|
|
|
bool use_only_the_last_commit_time_batch_for_recovery_ = false;
|
2020-04-30 23:21:05 +00:00
|
|
|
// Refer to
|
|
|
|
// TransactionOptions::skip_prepare
|
|
|
|
bool skip_prepare_ = false;
|
2017-11-02 00:23:52 +00:00
|
|
|
|
2017-08-07 23:07:40 +00:00
|
|
|
virtual Status PrepareInternal() = 0;
|
|
|
|
|
|
|
|
virtual Status CommitWithoutPrepareInternal() = 0;
|
|
|
|
|
2018-02-06 02:32:54 +00:00
|
|
|
// batch_cnt if non-zero is the number of sub-batches. A sub-batch is a batch
|
|
|
|
// with no duplicate keys. If zero, then the number of sub-batches is unknown.
|
|
|
|
virtual Status CommitBatchInternal(WriteBatch* batch,
|
|
|
|
size_t batch_cnt = 0) = 0;
|
2017-09-08 22:53:51 +00:00
|
|
|
|
2017-08-07 23:07:40 +00:00
|
|
|
virtual Status CommitInternal() = 0;
|
|
|
|
|
2017-10-03 02:46:42 +00:00
|
|
|
virtual Status RollbackInternal() = 0;
|
|
|
|
|
2018-07-24 07:09:18 +00:00
|
|
|
virtual void Initialize(const TransactionOptions& txn_options);
|
2017-08-03 15:46:47 +00:00
|
|
|
|
Replace tracked_keys with a new LockTracker interface in TransactionDB (#7013)
Summary:
We're going to support more locking protocols such as range lock in transaction.
However, in current design, `TransactionBase` has a member `tracked_keys` which assumes that point lock (lock a single key) is used, and is used in snapshot checking (isolation protocol). When using range lock, we may use read committed instead of snapshot checking as the isolation protocol.
The most significant usage scenarios of `tracked_keys` are:
1. pessimistic transaction uses it to track the locked keys, and unlock these keys when commit or rollback.
2. optimistic transaction does not lock keys upfront, it only tracks the lock intentions in tracked_keys, and do write conflict checking when commit.
3. each `SavePoint` tracks the keys that are locked since the `SavePoint`, `RollbackToSavePoint` or `PopSavePoint` relies on both the tracked keys in `SavePoint`s and `tracked_keys`.
Based on these scenarios, if we can abstract out a `LockTracker` interface to hold a set of tracked locks (can be keys or key ranges), and have methods that can be composed together to implement the scenarios, then `tracked_keys` can be an internal data structure of one implementation of `LockTracker`. See `utilities/transactions/lock/lock_tracker.h` for the detailed interface design, and `utilities/transactions/lock/point_lock_tracker.cc` for the implementation.
In the future, a `RangeLockTracker` can be implemented to track range locks without affecting other components.
After this PR, a clean interface for lock manager should be possible, and then ideally, we can have pluggable locking protocols.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7013
Test Plan: Run `transaction_test` and `optimistic_transaction_test`.
Reviewed By: ajkr
Differential Revision: D22163706
Pulled By: cheng-chang
fbshipit-source-id: f2860577b5334e31dd2994f5bc6d7c40d502b1b4
2020-08-06 19:36:48 +00:00
|
|
|
Status LockBatch(WriteBatch* batch, LockTracker* keys_to_unlock);
|
2017-08-03 15:46:47 +00:00
|
|
|
|
2015-08-21 22:47:21 +00:00
|
|
|
Status TryLock(ColumnFamilyHandle* column_family, const Slice& key,
|
2018-12-07 01:46:57 +00:00
|
|
|
bool read_only, bool exclusive, const bool do_validate = true,
|
|
|
|
const bool assume_tracked = false) override;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-08-03 15:46:47 +00:00
|
|
|
void Clear() override;
|
|
|
|
|
2017-08-06 00:17:48 +00:00
|
|
|
PessimisticTransactionDB* txn_db_impl_;
|
2016-04-18 18:15:50 +00:00
|
|
|
DBImpl* db_impl_;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-08-03 15:46:47 +00:00
|
|
|
// If non-zero, this transaction should not be committed after this time (in
|
|
|
|
// microseconds according to Env->NowMicros())
|
|
|
|
uint64_t expiration_time_;
|
|
|
|
|
|
|
|
private:
|
2017-11-02 01:56:25 +00:00
|
|
|
friend class TransactionTest_ValidateSnapshotTest_Test;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
// Used to create unique ids for transactions.
|
|
|
|
static std::atomic<TransactionID> txn_id_counter_;
|
|
|
|
|
|
|
|
// Unique ID for this transaction
|
2016-02-03 03:19:17 +00:00
|
|
|
TransactionID txn_id_;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2016-12-06 01:18:14 +00:00
|
|
|
// IDs for the transactions that are blocking the current transaction.
|
2016-09-28 00:43:06 +00:00
|
|
|
//
|
2016-12-06 01:18:14 +00:00
|
|
|
// empty if current transaction is not waiting.
|
|
|
|
autovector<TransactionID> waiting_txn_ids_;
|
2016-09-28 00:43:06 +00:00
|
|
|
|
|
|
|
// The following two represents the (cf, key) that a transaction is waiting
|
|
|
|
// on.
|
|
|
|
//
|
|
|
|
// If waiting_key_ is not null, then the pointer should always point to
|
|
|
|
// a valid string object. The reason is that it is only non-null when the
|
|
|
|
// transaction is blocked in the TransactionLockMgr::AcquireWithTimeout
|
|
|
|
// function. At that point, the key string object is one of the function
|
|
|
|
// parameters.
|
|
|
|
uint32_t waiting_cf_id_;
|
|
|
|
const std::string* waiting_key_;
|
|
|
|
|
2016-12-06 01:18:14 +00:00
|
|
|
// Mutex protecting waiting_txn_ids_, waiting_cf_id_ and waiting_key_.
|
2016-09-28 00:43:06 +00:00
|
|
|
mutable std::mutex wait_mutex_;
|
|
|
|
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
// Timeout in microseconds when locking a key or -1 if there is no timeout.
|
|
|
|
int64_t lock_timeout_;
|
|
|
|
|
2016-09-28 08:23:33 +00:00
|
|
|
// Whether to perform deadlock detection or not.
|
|
|
|
bool deadlock_detect_;
|
|
|
|
|
|
|
|
// Whether to perform deadlock detection or not.
|
|
|
|
int64_t deadlock_detect_depth_;
|
|
|
|
|
2018-09-10 23:45:59 +00:00
|
|
|
// Refer to TransactionOptions::skip_concurrency_control
|
|
|
|
bool skip_concurrency_control_;
|
|
|
|
|
2017-11-02 01:56:25 +00:00
|
|
|
virtual Status ValidateSnapshot(ColumnFamilyHandle* column_family,
|
2017-11-11 21:08:22 +00:00
|
|
|
const Slice& key,
|
|
|
|
SequenceNumber* tracked_at_seq);
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-08-03 15:46:47 +00:00
|
|
|
void UnlockGetForUpdate(ColumnFamilyHandle* column_family,
|
|
|
|
const Slice& key) override;
|
|
|
|
};
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-08-07 23:07:40 +00:00
|
|
|
class WriteCommittedTxn : public PessimisticTransaction {
|
2017-08-03 15:46:47 +00:00
|
|
|
public:
|
2017-08-07 23:07:40 +00:00
|
|
|
WriteCommittedTxn(TransactionDB* db, const WriteOptions& write_options,
|
2017-08-16 23:49:11 +00:00
|
|
|
const TransactionOptions& txn_options);
|
2019-09-12 01:07:12 +00:00
|
|
|
// No copying allowed
|
|
|
|
WriteCommittedTxn(const WriteCommittedTxn&) = delete;
|
|
|
|
void operator=(const WriteCommittedTxn&) = delete;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
2017-08-07 23:07:40 +00:00
|
|
|
virtual ~WriteCommittedTxn() {}
|
2017-08-03 15:46:47 +00:00
|
|
|
|
|
|
|
private:
|
2017-08-07 23:07:40 +00:00
|
|
|
Status PrepareInternal() override;
|
|
|
|
|
|
|
|
Status CommitWithoutPrepareInternal() override;
|
|
|
|
|
2018-02-06 02:32:54 +00:00
|
|
|
Status CommitBatchInternal(WriteBatch* batch, size_t batch_cnt) override;
|
2017-09-08 22:53:51 +00:00
|
|
|
|
2017-08-07 23:07:40 +00:00
|
|
|
Status CommitInternal() override;
|
|
|
|
|
2017-10-03 02:46:42 +00:00
|
|
|
Status RollbackInternal() override;
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
};
|
|
|
|
|
2020-02-20 20:07:53 +00:00
|
|
|
} // namespace ROCKSDB_NAMESPACE
|
Pessimistic Transactions
Summary:
Initial implementation of Pessimistic Transactions. This diff contains the api changes discussed in D38913. This diff is pretty large, so let me know if people would prefer to meet up to discuss it.
MyRocks folks: please take a look at the API in include/rocksdb/utilities/transaction[_db].h and let me know if you have any issues.
Also, you'll notice a couple of TODOs in the implementation of RollbackToSavePoint(). After chatting with Siying, I'm going to send out a separate diff for an alternate implementation of this feature that implements the rollback inside of WriteBatch/WriteBatchWithIndex. We can then decide which route is preferable.
Next, I'm planning on doing some perf testing and then integrating this diff into MongoRocks for further testing.
Test Plan: Unit tests, db_bench parallel testing.
Reviewers: igor, rven, sdong, yhchiang, yoshinorim
Reviewed By: sdong
Subscribers: hermanlee4, maykov, spetrunia, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40869
2015-05-26 00:37:33 +00:00
|
|
|
|
|
|
|
#endif // ROCKSDB_LITE
|