path: root/src/mongo/db/s/metadata_manager.cpp

/**
 *    Copyright (C) 2018-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
 *    <http://www.mongodb.com/licensing/server-side-public-license>.
 *
 *    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_LOG_DEFAULT_COMPONENT ::mongo::logger::LogComponent::kSharding

#include "mongo/platform/basic.h"

#include "mongo/db/s/metadata_manager.h"

#include <memory>

#include "mongo/base/string_data.h"
#include "mongo/bson/simple_bsonobj_comparator.h"
#include "mongo/bson/util/builder.h"
#include "mongo/db/query/internal_plans.h"
#include "mongo/db/range_arithmetic.h"
#include "mongo/db/s/range_deletion_util.h"
#include "mongo/db/s/sharding_runtime_d_params_gen.h"
#include "mongo/logv2/log.h"
#include "mongo/s/grid.h"
#include "mongo/util/assert_util.h"
#include "mongo/util/fail_point.h"
#include "mongo/util/log.h"
#include "mongo/util/time_support.h"

namespace mongo {
namespace {
using TaskExecutor = executor::TaskExecutor;
using CallbackArgs = TaskExecutor::CallbackArgs;
}  // namespace

class RangePreserver : public ScopedCollectionMetadata::Impl {
public:
    // Must be called locked with the MetadataManager's _managerLock
    RangePreserver(WithLock,
                   std::shared_ptr<MetadataManager> metadataManager,
                   std::shared_ptr<MetadataManager::CollectionMetadataTracker> metadataTracker)
        : _metadataManager(std::move(metadataManager)),
          _metadataTracker(std::move(metadataTracker)) {
        ++_metadataTracker->usageCounter;
    }

    ~RangePreserver() {
        stdx::lock_guard<Latch> managerLock(_metadataManager->_managerLock);

        invariant(_metadataTracker->usageCounter != 0);
        if (--_metadataTracker->usageCounter == 0) {
            // MetadataManager doesn't care which usageCounter went to zero. It just retires all
            // that are older than the oldest metadata still in use by queries (some start out at
            // zero, some go to zero but can't be expired yet).
            //
            // Note that new instances of ScopedCollectionMetadata may get attached to
            // _metadata.back(), so its usage count can increase from zero, unlike other reference
            // counts.
            _metadataManager->_retireExpiredMetadata(managerLock);
        }
    }

    // This will only ever refer to the active metadata, so CollectionMetadata should never be
    // boost::none
    const CollectionMetadata& get() {
        invariant(_metadataTracker->metadata);
        return _metadataTracker->metadata.get();
    }

private:
    std::shared_ptr<MetadataManager> _metadataManager;
    std::shared_ptr<MetadataManager::CollectionMetadataTracker> _metadataTracker;
};

MetadataManager::MetadataManager(ServiceContext* serviceContext,
                                 NamespaceString nss,
                                 std::shared_ptr<TaskExecutor> executor,
                                 CollectionMetadata initialMetadata)
    : _serviceContext(serviceContext),
      _nss(std::move(nss)),
      _collectionUuid(*initialMetadata.getChunkManager()->getUUID()),
      _executor(std::move(executor)),
      _receivingChunks(SimpleBSONObjComparator::kInstance.makeBSONObjIndexedMap<BSONObj>()) {
    _metadata.emplace_back(std::make_shared<CollectionMetadataTracker>(std::move(initialMetadata)));
}

ScopedCollectionMetadata MetadataManager::getActiveMetadata(
    const boost::optional<LogicalTime>& atClusterTime) {
    stdx::lock_guard<Latch> lg(_managerLock);

    auto activeMetadataTracker = _metadata.back();
    const auto& activeMetadata = activeMetadataTracker->metadata;

    // We don't keep routing history for unsharded collections, so if the collection is unsharded
    // just return the active metadata
    if (!atClusterTime || !activeMetadata->isSharded()) {
        return ScopedCollectionMetadata(std::make_shared<RangePreserver>(
            lg, shared_from_this(), std::move(activeMetadataTracker)));
    }

    auto chunkManager = activeMetadata->getChunkManager();
    auto chunkManagerAtClusterTime = std::make_shared<ChunkManager>(
        chunkManager->getRoutingHistory(), atClusterTime->asTimestamp());

    class MetadataAtTimestamp : public ScopedCollectionMetadata::Impl {
    public:
        MetadataAtTimestamp(CollectionMetadata metadata) : _metadata(std::move(metadata)) {}

        const CollectionMetadata& get() override {
            return _metadata;
        }

    private:
        CollectionMetadata _metadata;
    };

    return ScopedCollectionMetadata(std::make_shared<MetadataAtTimestamp>(
        CollectionMetadata(chunkManagerAtClusterTime, activeMetadata->shardId())));
}

size_t MetadataManager::numberOfMetadataSnapshots() const {
    stdx::lock_guard<Latch> lg(_managerLock);
    invariant(!_metadata.empty());
    return _metadata.size() - 1;
}

int MetadataManager::numberOfEmptyMetadataSnapshots() const {
    stdx::lock_guard<Latch> lg(_managerLock);

    int emptyMetadataSnapshots = 0;
    for (const auto& collMetadataTracker : _metadata) {
        if (!collMetadataTracker->metadata)
            emptyMetadataSnapshots++;
    }

    return emptyMetadataSnapshots;
}

void MetadataManager::setFilteringMetadata(CollectionMetadata remoteMetadata) {
    stdx::lock_guard<Latch> lg(_managerLock);
    invariant(!_metadata.empty());
    // The active metadata should always be available (not boost::none)
    invariant(_metadata.back()->metadata);
    const auto& activeMetadata = _metadata.back()->metadata.get();

    // We already have the same or newer version
    if (activeMetadata.getCollVersion().epoch() == remoteMetadata.getCollVersion().epoch() &&
        activeMetadata.getCollVersion() >= remoteMetadata.getCollVersion()) {
        LOGV2_DEBUG(21984,
                    1,
                    "Ignoring update of active metadata {activeMetadata_Basic} with an older "
                    "{remoteMetadata_Basic}",
                    "activeMetadata_Basic"_attr = activeMetadata.toStringBasic(),
                    "remoteMetadata_Basic"_attr = remoteMetadata.toStringBasic());
        return;
    }

    LOGV2(21985,
          "Updating metadata for collection {nss_ns} from {activeMetadata_Basic} to "
          "{remoteMetadata_Basic} due to version change",
          "nss_ns"_attr = _nss.ns(),
          "activeMetadata_Basic"_attr = activeMetadata.toStringBasic(),
          "remoteMetadata_Basic"_attr = remoteMetadata.toStringBasic());

    // Resolve any receiving chunks, which might have completed by now
    for (auto it = _receivingChunks.begin(); it != _receivingChunks.end();) {
        const ChunkRange receivingRange(it->first, it->second);

        if (!remoteMetadata.rangeOverlapsChunk(receivingRange)) {
            ++it;
            continue;
        }

        // The remote metadata contains a chunk we were earlier in the process of receiving, so we
        // deem it successfully received
        LOGV2_DEBUG(21986,
                    2,
                    "Verified chunk {receivingRange} for collection {nss_ns} has been migrated to "
                    "this shard earlier",
                    "receivingRange"_attr = redact(receivingRange.toString()),
                    "nss_ns"_attr = _nss.ns());

        _receivingChunks.erase(it);
        it = _receivingChunks.begin();
    }

    _setActiveMetadata(lg, std::move(remoteMetadata));
}

void MetadataManager::_setActiveMetadata(WithLock wl, CollectionMetadata newMetadata) {
    _metadata.emplace_back(std::make_shared<CollectionMetadataTracker>(std::move(newMetadata)));
    _retireExpiredMetadata(wl);
}

void MetadataManager::_retireExpiredMetadata(WithLock) {
    // Remove entries with a usage count of 0 from the front of _metadata, which may schedule
    // orphans for deletion. We cannot remove an entry from the middle of _metadata because a
    // previous entry (whose usageCount is not 0) could have a query that is actually still
    // accessing those documents.
    while (_metadata.size() > 1 && !_metadata.front()->usageCounter) {
        _metadata.pop_front();
    }

    // To avoid memory build up of ChunkManager objects, we can clear the CollectionMetadata object
    // in an entry when its usageCount is 0 as long as it is not the last item in _metadata (which
    // is the active metadata). If _metadata is empty, decrementing iter will be out of bounds, so
    // we must check that the size is > 1 as well.
    if (_metadata.size() > 1) {
        auto iter = _metadata.begin();
        while (iter != (--_metadata.end())) {
            if ((*iter)->usageCounter == 0) {
                (*iter)->metadata = boost::none;
            }
            ++iter;
        }
    }
}

void MetadataManager::toBSONPending(BSONArrayBuilder& bb) const {
    stdx::lock_guard<Latch> lg(_managerLock);

    for (auto it = _receivingChunks.begin(); it != _receivingChunks.end(); ++it) {
        BSONArrayBuilder pendingBB(bb.subarrayStart());
        pendingBB.append(it->first);
        pendingBB.append(it->second);
        pendingBB.done();
    }
}

void MetadataManager::append(BSONObjBuilder* builder) const {
    stdx::lock_guard<Latch> lg(_managerLock);

    BSONArrayBuilder arr(builder->subarrayStart("rangesToClean"));
    for (auto const& [range, _] : _rangesScheduledForDeletion) {
        BSONObjBuilder obj;
        range.append(&obj);
        arr.append(obj.done());
    }

    BSONArrayBuilder pcArr(builder->subarrayStart("pendingChunks"));
    for (const auto& entry : _receivingChunks) {
        BSONObjBuilder obj;
        ChunkRange r = ChunkRange(entry.first, entry.second);
        r.append(&obj);
        pcArr.append(obj.done());
    }
    pcArr.done();

    invariant(!_metadata.empty());

    BSONArrayBuilder amrArr(builder->subarrayStart("activeMetadataRanges"));
    for (const auto& entry : _metadata.back()->metadata->getChunks()) {
        BSONObjBuilder obj;
        ChunkRange r = ChunkRange(entry.first, entry.second);
        r.append(&obj);
        amrArr.append(obj.done());
    }
    amrArr.done();
}

SharedSemiFuture<void> MetadataManager::beginReceive(ChunkRange const& range) {
    stdx::lock_guard<Latch> lg(_managerLock);
    invariant(!_metadata.empty());

    if (_overlapsInUseChunk(lg, range)) {
        return Status{ErrorCodes::RangeOverlapConflict,
                      "Documents in target range may still be in use on the destination shard."};
    }

    _receivingChunks.emplace(range.getMin().getOwned(), range.getMax().getOwned());

    LOGV2(21987,
          "Scheduling deletion of any documents in {nss_ns} range {range} before migrating in a "
          "chunk covering the range",
          "nss_ns"_attr = _nss.ns(),
          "range"_attr = redact(range.toString()));

    return _submitRangeForDeletion(
        lg, SemiFuture<void>::makeReady(), range, Seconds(orphanCleanupDelaySecs.load()));
}

void MetadataManager::forgetReceive(ChunkRange const& range) {
    stdx::lock_guard<Latch> lg(_managerLock);
    invariant(!_metadata.empty());

    // This is potentially a partially received chunk, which needs to be cleaned up. We know none
    // of these documents are in use, so they can go straight to the deletion queue.
    LOGV2(21988,
          "Abandoning in-migration of {nss_ns} range {range}; scheduling deletion of any documents "
          "already copied",
          "nss_ns"_attr = _nss.ns(),
          "range"_attr = range);

    invariant(!_overlapsInUseChunk(lg, range));

    auto it = _receivingChunks.find(range.getMin());
    invariant(it != _receivingChunks.end());
    _receivingChunks.erase(it);

    std::ignore = _submitRangeForDeletion(lg, SemiFuture<void>::makeReady(), range, Seconds(0));
}

SharedSemiFuture<void> MetadataManager::cleanUpRange(ChunkRange const& range,
                                                     bool shouldDelayBeforeDeletion) {
    stdx::lock_guard<Latch> lg(_managerLock);
    invariant(!_metadata.empty());

    auto* const activeMetadata = _metadata.back().get();
    auto* const overlapMetadata = _findNewestOverlappingMetadata(lg, range);

    if (overlapMetadata == activeMetadata) {
        return Status{ErrorCodes::RangeOverlapConflict,
                      str::stream() << "Requested deletion range overlaps a live shard chunk"};
    }

    if (rangeMapOverlaps(_receivingChunks, range.getMin(), range.getMax())) {
        return Status{ErrorCodes::RangeOverlapConflict,
                      str::stream() << "Requested deletion range overlaps a chunk being"
                                       " migrated in"};
    }

    auto delayForActiveQueriesOnSecondariesToComplete =
        shouldDelayBeforeDeletion ? Seconds(orphanCleanupDelaySecs.load()) : Seconds(0);

    if (overlapMetadata) {
        LOGV2(21989,
              "Deletion of {nss_ns} range {range} will be scheduled after all possibly dependent "
              "queries finish",
              "nss_ns"_attr = _nss.ns(),
              "range"_attr = redact(range.toString()));
        ++overlapMetadata->numContingentRangeDeletionTasks;
        // Schedule the range for deletion once the overlapping metadata object is destroyed
        // (meaning no more queries can be using the range) and obtain a future which will be
        // signaled when deletion is complete.
        return _submitRangeForDeletion(lg,
                                       overlapMetadata->onDestructionPromise.getFuture().semi(),
                                       range,
                                       delayForActiveQueriesOnSecondariesToComplete);
    } else {
        // No running queries can depend on this range, so queue it for deletion immediately.
        LOGV2(21990,
              "Scheduling deletion of {nss_ns} range {range}",
              "nss_ns"_attr = _nss.ns(),
              "range"_attr = redact(range.toString()));

        return _submitRangeForDeletion(
            lg, SemiFuture<void>::makeReady(), range, delayForActiveQueriesOnSecondariesToComplete);
    }
}

size_t MetadataManager::numberOfRangesToCleanStillInUse() const {
    stdx::lock_guard<Latch> lg(_managerLock);
    size_t count = 0;
    for (auto& tracker : _metadata) {
        count += tracker->numContingentRangeDeletionTasks;
    }
    return count;
}

size_t MetadataManager::numberOfRangesToClean() const {
    auto rangesToCleanInUse = numberOfRangesToCleanStillInUse();
    stdx::lock_guard<Latch> lg(_managerLock);
    return _rangesScheduledForDeletion.size() - rangesToCleanInUse;
}

boost::optional<SharedSemiFuture<void>> MetadataManager::trackOrphanedDataCleanup(
    ChunkRange const& range) const {
    stdx::lock_guard<Latch> lg(_managerLock);
    for (const auto& [orphanRange, deletionComplete] : _rangesScheduledForDeletion) {
        if (orphanRange.overlapWith(range)) {
            return deletionComplete;
        }
    }

    return boost::none;
}

auto MetadataManager::_findNewestOverlappingMetadata(WithLock, ChunkRange const& range)
    -> CollectionMetadataTracker* {
    invariant(!_metadata.empty());

    auto it = _metadata.rbegin();
    if ((*it)->metadata && (*it)->metadata->rangeOverlapsChunk(range)) {
        return (*it).get();
    }

    ++it;
    for (; it != _metadata.rend(); ++it) {
        auto& tracker = *it;
        if (tracker->usageCounter && tracker->metadata &&
            tracker->metadata->rangeOverlapsChunk(range)) {
            return tracker.get();
        }
    }

    return nullptr;
}

bool MetadataManager::_overlapsInUseChunk(WithLock lk, ChunkRange const& range) {
    auto* cm = _findNewestOverlappingMetadata(lk, range);
    return (cm != nullptr);
}

boost::optional<ChunkRange> MetadataManager::getNextOrphanRange(BSONObj const& from) const {
    stdx::lock_guard<Latch> lg(_managerLock);
    invariant(!_metadata.empty());
    return _metadata.back()->metadata->getNextOrphanRange(_receivingChunks, from);
}

SharedSemiFuture<void> MetadataManager::_submitRangeForDeletion(
    const WithLock&,
    SemiFuture<void> waitForActiveQueriesToComplete,
    const ChunkRange& range,
    Seconds delayForActiveQueriesOnSecondariesToComplete) {

    int maxToDelete = rangeDeleterBatchSize.load();
    if (maxToDelete <= 0) {
        maxToDelete = std::max(int(internalQueryExecYieldIterations.load()), 1);
    }

    auto cleanupComplete =
        removeDocumentsInRange(_executor,
                               std::move(waitForActiveQueriesToComplete),
                               _nss,
                               *_metadata.back()->metadata->getChunkManager()->getUUID(),
                               _metadata.back()->metadata->getKeyPattern().getOwned(),
                               range,
                               maxToDelete,
                               delayForActiveQueriesOnSecondariesToComplete,
                               Milliseconds(rangeDeleterBatchDelayMS.load()));

    _rangesScheduledForDeletion.emplace_front(range, cleanupComplete);
    // Attach a continuation so that once the range has been deleted, we will remove the deletion
    // from the _rangesScheduledForDeletion.  std::list iterators are never invalidated, which
    // allows us to save the iterator pointing to the newly added element for use later when
    // deleting it.
    cleanupComplete.thenRunOn(_executor).getAsync(
        [self = shared_from_this(), it = _rangesScheduledForDeletion.begin()](Status s) {
            stdx::lock_guard<Latch> lg(self->_managerLock);
            self->_rangesScheduledForDeletion.erase(it);
        });
    return cleanupComplete;
}
}  // namespace mongo