/** * Copyright (C) 2020-present MongoDB, Inc. * * This program is free software: you can redistribute it and/or modify * it under the terms of the Server Side Public License, version 1, * as published by MongoDB, Inc. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * Server Side Public License for more details. * * You should have received a copy of the Server Side Public License * along with this program. If not, see * . * * As a special exception, the copyright holders give permission to link the * code of portions of this program with the OpenSSL library under certain * conditions as described in each individual source file and distribute * linked combinations including the program with the OpenSSL library. You * must comply with the Server Side Public License in all respects for * all of the code used other than as permitted herein. If you modify file(s) * with this exception, you may extend this exception to your version of the * file(s), but you are not obligated to do so. If you do not wish to do so, * delete this exception statement from your version. If you delete this * exception statement from all source files in the program, then also delete * it in the license file. */ #define MONGO_LOGV2_DEFAULT_COMPONENT ::mongo::logv2::LogComponent::kResharding #include "mongo/platform/basic.h" #include "mongo/db/s/resharding/resharding_oplog_applier.h" #include "mongo/db/persistent_task_store.h" #include "mongo/db/s/resharding/resharding_data_copy_util.h" #include "mongo/db/s/resharding/resharding_donor_oplog_iterator.h" #include "mongo/db/s/resharding/resharding_future_util.h" #include "mongo/db/s/resharding/resharding_metrics.h" #include "mongo/db/s/resharding/resharding_util.h" #include "mongo/logv2/log.h" #include "mongo/s/sharding_feature_flags_gen.h" #include "mongo/util/assert_util.h" #include "mongo/util/fail_point.h" #include "mongo/util/uuid.h" namespace mongo { namespace { MONGO_FAIL_POINT_DEFINE(reshardingApplyOplogBatchTwice); } ReshardingOplogApplier::ReshardingOplogApplier( std::unique_ptr env, ReshardingSourceId sourceId, NamespaceString outputNss, std::vector allStashNss, size_t myStashIdx, ChunkManager sourceChunkMgr, std::unique_ptr oplogIterator) : _env(std::move(env)), _sourceId(std::move(sourceId)), _batchPreparer{CollatorInterface::cloneCollator(sourceChunkMgr.getDefaultCollator())}, _crudApplication{std::move(outputNss), std::move(allStashNss), myStashIdx, _sourceId.getShardId(), std::move(sourceChunkMgr), _env->metrics(), _env->metricsNew()}, _sessionApplication{}, _batchApplier{_crudApplication, _sessionApplication}, _oplogIter(std::move(oplogIterator)) {} SemiFuture ReshardingOplogApplier::_applyBatch( std::shared_ptr executor, CancellationToken cancelToken, CancelableOperationContextFactory factory) { Timer latencyTimer; auto crudWriterVectors = _batchPreparer.makeCrudOpWriterVectors( _currentBatchToApply, _currentDerivedOpsForCrudWriters); CancellationSource errorSource(cancelToken); std::vector> batchApplierFutures; // Use `2 * crudWriterVectors.size()` because sessionWriterVectors.size() is very likely equal // to crudWriterVectors.size(). Calling ReshardingOplogBatchApplier::applyBatch() first // though allows CRUD application to be concurrent with preparing the writer vectors for session // application in addition to being concurrent with session application itself. batchApplierFutures.reserve(2 * crudWriterVectors.size()); for (auto&& writer : crudWriterVectors) { if (!writer.empty()) { batchApplierFutures.emplace_back( _batchApplier .applyBatch(std::move(writer), executor, errorSource.token(), factory) .share()); } } auto sessionWriterVectors = _batchPreparer.makeSessionOpWriterVectors( _currentBatchToApply, _currentDerivedOpsForSessionWriters); batchApplierFutures.reserve(crudWriterVectors.size() + sessionWriterVectors.size()); for (auto&& writer : sessionWriterVectors) { if (!writer.empty()) { batchApplierFutures.emplace_back( _batchApplier .applyBatch(std::move(writer), executor, errorSource.token(), factory) .share()); } } return resharding::cancelWhenAnyErrorThenQuiesce(batchApplierFutures, executor, errorSource) .onError([](Status status) { LOGV2_ERROR( 5012004, "Failed to apply operation in resharding", "error"_attr = redact(status)); return status; }) .onCompletion([this, latencyTimer](Status status) { _env->metrics()->onOplogApplierApplyBatch( duration_cast(latencyTimer.elapsed())); return status; }) .semi(); } SemiFuture ReshardingOplogApplier::run( std::shared_ptr executor, std::shared_ptr cleanupExecutor, CancellationToken cancelToken, CancelableOperationContextFactory factory) { struct ChainContext { std::unique_ptr oplogIter; }; auto chainCtx = std::make_shared(); chainCtx->oplogIter = std::move(_oplogIter); return AsyncTry([this, chainCtx, executor, cancelToken, factory] { return chainCtx->oplogIter->getNextBatch(executor, cancelToken, factory) .thenRunOn(executor) .then([this, executor, cancelToken, factory](OplogBatch batch) { LOGV2_DEBUG(5391002, 3, "Starting batch", "batchSize"_attr = batch.size()); _currentBatchToApply = std::move(batch); return _applyBatch(executor, cancelToken, factory); }) .then([this, executor, cancelToken, factory] { if (MONGO_unlikely(reshardingApplyOplogBatchTwice.shouldFail())) { LOGV2(5687600, "reshardingApplyOplogBatchTwice failpoint enabled, applying batch " "a second time", "batchSize"_attr = _currentBatchToApply.size()); _currentDerivedOpsForCrudWriters.clear(); _currentDerivedOpsForSessionWriters.clear(); return _applyBatch(executor, cancelToken, factory); } return SemiFuture(); }) .then([this, factory] { if (_currentBatchToApply.empty()) { // Increment the number of entries applied by 1 in order to account for // the final oplog entry. _env->metrics()->onOplogEntriesApplied(1); return false; } auto opCtx = factory.makeOperationContext(&cc()); _clearAppliedOpsAndStoreProgress(opCtx.get()); return true; }); }) .until([](const StatusWith& swMoreToApply) { return !swMoreToApply.isOK() || !swMoreToApply.getValue(); }) .on(executor, cancelToken) .ignoreValue() .thenRunOn(std::move(cleanupExecutor)) // It is unsafe to capture `this` once the task is running on the cleanupExecutor because // RecipientStateMachine, along with its ReshardingOplogApplier member, may have already // been destructed. .onCompletion([chainCtx](Status status) { if (chainCtx->oplogIter) { // Use a separate Client to make a better effort of calling dispose() even when the // CancellationToken has been canceled. auto client = cc().getServiceContext()->makeClient("ReshardingOplogApplierCleanupClient"); AlternativeClientRegion acr(client); auto opCtx = cc().makeOperationContext(); chainCtx->oplogIter->dispose(opCtx.get()); chainCtx->oplogIter.reset(); } return status; }) .semi(); } boost::optional ReshardingOplogApplier::checkStoredProgress( OperationContext* opCtx, const ReshardingSourceId& id) { DBDirectClient client(opCtx); auto doc = client.findOne( NamespaceString::kReshardingApplierProgressNamespace, BSON(ReshardingOplogApplierProgress::kOplogSourceIdFieldName << id.toBSON())); if (doc.isEmpty()) { return boost::none; } IDLParserErrorContext ctx("ReshardingOplogApplierProgress"); return ReshardingOplogApplierProgress::parse(ctx, doc); } void ReshardingOplogApplier::_clearAppliedOpsAndStoreProgress(OperationContext* opCtx) { const auto& lastOplog = _currentBatchToApply.back(); auto oplogId = ReshardingDonorOplogId::parse({"ReshardingOplogApplier::_clearAppliedOpsAndStoreProgress"}, lastOplog.get_id()->getDocument().toBson()); PersistentTaskStore store( NamespaceString::kReshardingApplierProgressNamespace); BSONObjBuilder builder; builder.append("$set", BSON(ReshardingOplogApplierProgress::kProgressFieldName << oplogId.toBSON())); builder.append("$inc", BSON(ReshardingOplogApplierProgress::kNumEntriesAppliedFieldName << static_cast(_currentBatchToApply.size()))); store.upsert( opCtx, BSON(ReshardingOplogApplierProgress::kOplogSourceIdFieldName << _sourceId.toBSON()), builder.obj()); _env->metrics()->onOplogEntriesApplied(_currentBatchToApply.size()); if (feature_flags::gFeatureFlagShardingDataTransformMetrics.isEnabledAndIgnoreFCV()) { _env->metricsNew()->onOplogEntriesApplied(_currentBatchToApply.size()); } _currentBatchToApply.clear(); _currentDerivedOpsForCrudWriters.clear(); _currentDerivedOpsForSessionWriters.clear(); } NamespaceString ReshardingOplogApplier::ensureStashCollectionExists( OperationContext* opCtx, const UUID& existingUUID, const ShardId& donorShardId, const CollectionOptions& options) { auto nss = getLocalConflictStashNamespace(existingUUID, donorShardId); resharding::data_copy::ensureCollectionExists(opCtx, nss, options); return nss; } } // namespace mongo