SERVER-35612 Implement ChunkWritesTracker with unit tests, make ChunkInfo use it

11 files changed, 740 insertions, 35 deletions

diff --git a/src/mongo/db/s/shard_server_op_observer.cpp b/src/mongo/db/s/shard_server_op_observer.cpp
index 9b8ba53ae89..20508022dbf 100644
--- a/src/mongo/db/s/shard_server_op_observer.cpp
+++ b/src/mongo/db/s/shard_server_op_observer.cpp
@@ -145,11 +145,7 @@ bool shouldSplitChunk(OperationContext* opCtx,
     const auto balancerConfig = Grid::get(opCtx)->getBalancerConfiguration();
     invariant(balancerConfig);
 
-    const auto& keyPattern = shardKeyPattern.getKeyPattern();
-    const bool minIsInf = (0 == keyPattern.globalMin().woCompare(chunk.getMin()));
-    const bool maxIsInf = (0 == keyPattern.globalMax().woCompare(chunk.getMax()));
-
-    return chunk.shouldSplit(balancerConfig->getMaxChunkSizeBytes(), minIsInf, maxIsInf);
+    return chunk.shouldSplit(balancerConfig->getMaxChunkSizeBytes());
 }
 
 /**
diff --git a/src/mongo/s/SConscript b/src/mongo/s/SConscript
index bfce3ffc687..bda3163ced7 100644
--- a/src/mongo/s/SConscript
+++ b/src/mongo/s/SConscript
@@ -64,6 +64,7 @@ env.Library(
         '$BUILD_DIR/mongo/db/storage/key_string',
         '$BUILD_DIR/mongo/db/update/update_common',
         '$BUILD_DIR/mongo/util/concurrency/ticketholder',
+        'chunk_writes_tracker',
         'common_s',
     ],
 )
@@ -363,6 +364,46 @@ env.CppUnitTest(
 )
 
 env.Library(
+    target='chunk_writes_tracker',
+    source=[
+        'chunk_writes_tracker.cpp',
+    ],
+    LIBDEPS=[
+        '$BUILD_DIR/mongo/base',
+    ],
+)
+
+env.CppUnitTest(
+    target='chunk_writes_tracker_test',
+    source=[
+        'chunk_writes_tracker_test.cpp',
+    ],
+    LIBDEPS=[
+        'chunk_writes_tracker',
+    ]
+)
+
+env.Library(
+    target='chunk_split_state_driver',
+    source=[
+        'chunk_split_state_driver.cpp',
+    ],
+    LIBDEPS=[
+        'chunk_writes_tracker'
+    ],
+)
+
+env.CppUnitTest(
+    target='chunk_split_state_driver_test',
+    source=[
+        'chunk_split_state_driver_test.cpp',
+    ],
+    LIBDEPS=[
+        'chunk_split_state_driver',
+    ]
+)
+
+env.Library(
     target='catalog_cache_test_fixture',
     source=[
         'catalog_cache_test_fixture.cpp',
diff --git a/src/mongo/s/chunk.cpp b/src/mongo/s/chunk.cpp
index b32c6d375f4..2bc6c8748ac 100644
--- a/src/mongo/s/chunk.cpp
+++ b/src/mongo/s/chunk.cpp
@@ -33,6 +33,7 @@
 #include "mongo/s/chunk.h"
 
 #include "mongo/platform/random.h"
+#include "mongo/s/chunk_writes_tracker.h"
 #include "mongo/util/mongoutils/str.h"
 
 namespace mongo {
@@ -43,7 +44,7 @@ PseudoRandom prng(static_cast<int64_t>(time(0)));
 
 // Assume user has 64MB chunkSize setting. It is ok if this assumption is wrong since it is only
 // a heuristic.
-const int64_t kMaxDataWritten = 64 / Chunk::kSplitTestFactor;
+const int64_t kMaxDataWritten = 64 / ChunkWritesTracker::kSplitTestFactor;
 
 /**
  * Generates a random value for _dataWritten so that a mongos restart wouldn't cause delay in
@@ -61,11 +62,12 @@ ChunkInfo::ChunkInfo(const ChunkType& from)
       _lastmod(from.getVersion()),
       _history(from.getHistory()),
       _jumbo(from.getJumbo()),
-      _dataWritten(mkDataWritten()) {
+      _writesTracker(std::make_shared<ChunkWritesTracker>()) {
     invariant(from.validate());
     if (!_history.empty()) {
         invariant(_shardId == _history.front().getShard());
     }
+    _writesTracker->addBytesWritten(mkDataWritten());
 }
 
 const ShardId& ChunkInfo::getShardIdAt(const boost::optional<Timestamp>& ts) const {
@@ -97,27 +99,20 @@ bool ChunkInfo::containsKey(const BSONObj& shardKey) const {
 }
 
 uint64_t ChunkInfo::getBytesWritten() const {
-    return _dataWritten;
+    return _writesTracker->getBytesWritten();
 }
 
-uint64_t ChunkInfo::addBytesWritten(uint64_t bytesWrittenIncrement) {
-    _dataWritten += bytesWrittenIncrement;
-    return _dataWritten;
+void ChunkInfo::addBytesWritten(uint64_t bytesWrittenIncrement) {
+    _writesTracker->addBytesWritten(bytesWrittenIncrement);
 }
 
 void ChunkInfo::clearBytesWritten() {
-    _dataWritten = 0;
+    _writesTracker->clearBytesWritten();
 }
 
-bool ChunkInfo::shouldSplit(uint64_t desiredChunkSize, bool minIsInf, bool maxIsInf) const {
-    // If this chunk is at either end of the range, trigger auto-split at 10% less data written in
-    // order to trigger the top-chunk optimization.
-    const uint64_t splitThreshold = (minIsInf || maxIsInf)
-        ? static_cast<uint64_t>((double)desiredChunkSize * 0.9)
-        : desiredChunkSize;
-
+bool ChunkInfo::shouldSplit(uint64_t maxChunkSize) const {
     // Check if there are enough estimated bytes written to warrant a split
-    return _dataWritten >= splitThreshold / Chunk::kSplitTestFactor;
+    return _writesTracker->shouldSplit(maxChunkSize);
 }
 
 std::string ChunkInfo::toString() const {
diff --git a/src/mongo/s/chunk.h b/src/mongo/s/chunk.h
index 34a2e02b57d..4e26d006402 100644
--- a/src/mongo/s/chunk.h
+++ b/src/mongo/s/chunk.h
@@ -35,6 +35,7 @@
 namespace mongo {
 
 class BSONObj;
+class ChunkWritesTracker;
 
 /**
  * Represents a cache entry for a single Chunk. Owned by a RoutingTableHistory.
@@ -80,10 +81,10 @@ public:
      * Get/increment/set the estimation of how much data was written for this chunk.
      */
     uint64_t getBytesWritten() const;
-    uint64_t addBytesWritten(uint64_t bytesWrittenIncrement);
+    void addBytesWritten(uint64_t bytesWrittenIncrement);
     void clearBytesWritten();
 
-    bool shouldSplit(uint64_t desiredChunkSize, bool minIsInf, bool maxIsInf) const;
+    bool shouldSplit(uint64_t desiredChunkSize) const;
 
     /**
      * Marks this chunk as jumbo. Only moves from false to true once and is used by the balancer.
@@ -103,15 +104,14 @@ private:
     // split
     mutable bool _jumbo;
 
-    // Statistics for the approximate data written to this chunk
-    mutable uint64_t _dataWritten;
+    // Used for tracking writes to this chunk, to estimate its size for the autosplitter. Since
+    // ChunkInfo obejcts are always treated as const, and this contains metadata about the chunk
+    // that needs to change, it's okay (and necessary) to mark it mutable
+    mutable std::shared_ptr<ChunkWritesTracker> _writesTracker;
 };
 
 class Chunk {
 public:
-    // Test whether we should split once data * kSplitTestFactor > chunkSize (approximately)
-    static const uint64_t kSplitTestFactor = 5;
-
     Chunk(ChunkInfo& chunkInfo, const boost::optional<Timestamp>& atClusterTime)
         : _chunkInfo(chunkInfo), _atClusterTime(atClusterTime) {}
 
@@ -160,15 +160,15 @@ public:
     uint64_t getBytesWritten() const {
         return _chunkInfo.getBytesWritten();
     }
-    uint64_t addBytesWritten(uint64_t bytesWrittenIncrement) {
-        return _chunkInfo.addBytesWritten(bytesWrittenIncrement);
+    void addBytesWritten(uint64_t bytesWrittenIncrement) {
+        _chunkInfo.addBytesWritten(bytesWrittenIncrement);
     }
     void clearBytesWritten() {
         _chunkInfo.clearBytesWritten();
     }
 
-    bool shouldSplit(uint64_t desiredChunkSize, bool minIsInf, bool maxIsInf) const {
-        return _chunkInfo.shouldSplit(desiredChunkSize, minIsInf, maxIsInf);
+    bool shouldSplit(uint64_t desiredChunkSize) const {
+        return _chunkInfo.shouldSplit(desiredChunkSize);
     }
 
     /**
diff --git a/src/mongo/s/chunk_split_state_driver.cpp b/src/mongo/s/chunk_split_state_driver.cpp
new file mode 100644
index 00000000000..707033d6961
--- /dev/null
+++ b/src/mongo/s/chunk_split_state_driver.cpp
@@ -0,0 +1,89 @@
+/**
+ *    Copyright (C) 2018 MongoDB Inc.
+ *
+ *    This program is free software: you can redistribute it and/or  modify
+ *    it under the terms of the GNU Affero General Public License, version 3,
+ *    as published by the Free Software Foundation.
+ *
+ *    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
+ *    GNU Affero General Public License for more details.
+ *
+ *    You should have received a copy of the GNU Affero General Public License
+ *    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ *    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 GNU Affero General Public License in all respects for
+ *    all of the code used source 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/s/chunk_split_state_driver.h"
+
+#include "mongo/util/assert_util.h"
+
+namespace mongo {
+
+boost::optional<ChunkSplitStateDriver> ChunkSplitStateDriver::tryInitiateSplit(
+    std::shared_ptr<ChunkWritesTracker> writesTracker) {
+    bool acquiredSplitLock = writesTracker->acquireSplitLock();
+    return acquiredSplitLock
+        ? boost::optional<ChunkSplitStateDriver>(ChunkSplitStateDriver(writesTracker))
+        : boost::none;
+}
+
+ChunkSplitStateDriver::ChunkSplitStateDriver(ChunkSplitStateDriver&& source) {
+    _writesTracker = source._writesTracker;
+    source._writesTracker.reset();
+    _splitState = source._splitState;
+    // Ensure that the source driver does not cancel the ongoing split
+    source._splitState = SplitState::kNotSplitting;
+}
+
+ChunkSplitStateDriver::~ChunkSplitStateDriver() {
+    if (_splitState == SplitState::kSplitInProgress || _splitState == SplitState::kSplitPrepared) {
+        cancelSplit();
+    }
+}
+
+void ChunkSplitStateDriver::prepareSplit() {
+    invariant(_splitState == SplitState::kSplitInProgress);
+    _splitState = SplitState::kSplitPrepared;
+
+    _stashedBytesWritten = _getWritesTracker()->getBytesWritten();
+    _getWritesTracker()->clearBytesWritten();
+}
+
+void ChunkSplitStateDriver::commitSplit() {
+    invariant(_splitState == SplitState::kSplitPrepared);
+    _splitState = SplitState::kSplitCommitted;
+}
+
+void ChunkSplitStateDriver::cancelSplit() {
+    invariant(_splitState == SplitState::kSplitInProgress ||
+              _splitState == SplitState::kSplitPrepared);
+    _splitState = SplitState::kNotSplitting;
+
+    _getWritesTracker()->releaseSplitLock();
+    _getWritesTracker()->addBytesWritten(_stashedBytesWritten);
+}
+
+std::shared_ptr<ChunkWritesTracker> ChunkSplitStateDriver::_getWritesTracker() {
+    auto wt = _writesTracker.lock();
+    invariant(wt, "ChunkWritesTracker was destructed before ChunkSplitStateDriver");
+    return wt;
+}
+
+}  // namespace mongo
diff --git a/src/mongo/s/chunk_split_state_driver.h b/src/mongo/s/chunk_split_state_driver.h
new file mode 100644
index 00000000000..427f3cda056
--- /dev/null
+++ b/src/mongo/s/chunk_split_state_driver.h
@@ -0,0 +1,131 @@
+/**
+ *    Copyright (C) 2018 MongoDB Inc.
+ *
+ *    This program is free software: you can redistribute it and/or  modify
+ *    it under the terms of the GNU Affero General Public License, version 3,
+ *    as published by the Free Software Foundation.
+ *
+ *    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
+ *    GNU Affero General Public License for more details.
+ *
+ *    You should have received a copy of the GNU Affero General Public License
+ *    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ *    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 GNU Affero General 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.
+ */
+
+#pragma once
+
+
+#include <boost/optional.hpp>
+#include <memory>
+
+#include "mongo/base/disallow_copying.h"
+#include "mongo/s/chunk_writes_tracker.h"
+
+
+namespace mongo {
+
+/**
+ * Drives state transitions and the status of a ChunkWritesTracker for a chunk
+ * while it is splitting. Supports cancelability so that we don't lose
+ * information from the writes tracker if the split errors. Should be created
+ * when it is decided that a split should be performed and then passed along to
+ * the ChunkSplitter which will drive these state changes.
+ */
+class ChunkSplitStateDriver {
+    MONGO_DISALLOW_COPYING(ChunkSplitStateDriver);
+
+public:
+    /**
+     * Tries to lock the writesTracker for splitting, and if it succeeds,
+     * constructs and returns a ChunkSplitStateDriver object. If it fails due to the
+     * writesTracker already being locked, returns boost::none.
+     */
+    static boost::optional<ChunkSplitStateDriver> tryInitiateSplit(
+        std::shared_ptr<ChunkWritesTracker> writesTracker);
+
+    /**
+     * Sets the other pointer back to default initialized state so that it
+     * doesn't try to cancel any ongoing split in its destructor. This
+     * constructor is required for boost::optional.
+     */
+    ChunkSplitStateDriver(ChunkSplitStateDriver&& other);
+
+    /**
+     * Not needed.
+     */
+    ChunkSplitStateDriver& operator=(ChunkSplitStateDriver&&) = delete;
+
+    /**
+     * If there's an ongoing split, cancels it. Otherwise does nothing.
+     */
+    ~ChunkSplitStateDriver();
+
+    /**
+     * Clears the current bytes written, but stashes them in a variable in case
+     * the split is later canceled.
+     */
+    void prepareSplit();
+
+    /**
+     * Marks the split as committed, which means that shouldSplit will
+     * never again return true.
+     */
+    void commitSplit();
+
+    /**
+     * Marks the split state of an in-progress split back to kNotSplitting.
+     * If a split has already been prepared, resets the byte counter to what it
+     * was prior to prepare plus any new bytes that have been written.
+     */
+    void cancelSplit();
+
+private:
+    /**
+     * Should only be used by tryInitiateSplit
+     */
+    ChunkSplitStateDriver(std::shared_ptr<ChunkWritesTracker> writesTracker)
+        : _writesTracker(writesTracker), _splitState(SplitState::kSplitInProgress) {}
+
+    /**
+     * Pointer to the writes tracker object for which we're driving the split
+     */
+    std::weak_ptr<ChunkWritesTracker> _writesTracker;
+
+    /**
+     * Carries over result from prepare into cancelSplit.
+     */
+    uint64_t _stashedBytesWritten{0};
+
+    /**
+     * The current state of the chunk with respect to its progress being split.
+     */
+    enum class SplitState {
+        kNotSplitting,
+        kSplitInProgress,
+        kSplitPrepared,
+        kSplitCommitted,
+    } _splitState{SplitState::kNotSplitting};
+
+
+    /**
+     * Returns the ChunkWritesTracker whose state this driver is controlling,
+     * checking to make sure it has not yet been destroyed.
+     */
+    std::shared_ptr<ChunkWritesTracker> _getWritesTracker();
+};
+
+}  // namespace mongo
diff --git a/src/mongo/s/chunk_split_state_driver_test.cpp b/src/mongo/s/chunk_split_state_driver_test.cpp
new file mode 100644
index 00000000000..89042b3dbec
--- /dev/null
+++ b/src/mongo/s/chunk_split_state_driver_test.cpp
@@ -0,0 +1,142 @@
+/**
+ * Copyright (C) 2018 MongoDB Inc.
+ *
+ * This program is free software: you can redistribute it and/or  modify
+ * it under the terms of the GNU Affero General Public License, version 3,
+ * as published by the Free Software Foundation.
+ *
+ * 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
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ * 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 GNU Affero General 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.
+ */
+
+#include "mongo/platform/basic.h"
+
+#include "mongo/s/chunk_split_state_driver.h"
+
+#include "mongo/unittest/death_test.h"
+#include "mongo/unittest/unittest.h"
+
+
+namespace mongo {
+
+class ChunkSplitStateDriverTest : public unittest::Test {
+public:
+    // Add bytes to write tracker and create the ChunkSplitStateDriver object
+    // to test, which starts the split on the writes tracker
+    void setUp() override {
+        _writesTracker = std::make_shared<ChunkWritesTracker>();
+        uint64_t bytesToAdd{4};
+        _writesTracker->addBytesWritten(bytesToAdd);
+        _splitDriver = std::make_unique<boost::optional<ChunkSplitStateDriver>>(
+            ChunkSplitStateDriver::tryInitiateSplit(_writesTracker));
+    }
+
+    void tearDown() override {
+        _splitDriver.reset();
+        _writesTracker.reset();
+    }
+
+    ChunkWritesTracker& writesTracker() {
+        return *_writesTracker;
+    }
+
+    boost::optional<ChunkSplitStateDriver>& splitDriver() {
+        return *_splitDriver;
+    }
+
+private:
+    std::shared_ptr<ChunkWritesTracker> _writesTracker;
+    std::unique_ptr<boost::optional<ChunkSplitStateDriver>> _splitDriver;
+};
+
+TEST_F(ChunkSplitStateDriverTest, InitiateSplitLeavesBytesWrittenUnchanged) {
+    auto bytesInTracker = writesTracker().getBytesWritten();
+    ASSERT_TRUE(splitDriver());
+    ASSERT_EQ(writesTracker().getBytesWritten(), bytesInTracker);
+}
+
+TEST_F(ChunkSplitStateDriverTest, PrepareSplitClearsBytesWritten) {
+    splitDriver()->prepareSplit();
+    ASSERT_EQ(writesTracker().getBytesWritten(), 0ull);
+}
+
+TEST_F(ChunkSplitStateDriverTest, PrepareSplitFollowedByCancelSplitRestoresBytesWritten) {
+    auto bytesInTracker = writesTracker().getBytesWritten();
+    splitDriver()->prepareSplit();
+    splitDriver()->cancelSplit();
+    ASSERT_EQ(writesTracker().getBytesWritten(), bytesInTracker);
+}
+
+TEST_F(ChunkSplitStateDriverTest,
+       PrepareSplitThenAddBytesThenCancelSplitRestoresOldBytesWrittenPlusNewBytesWritten) {
+    auto bytesInTracker = writesTracker().getBytesWritten();
+    splitDriver()->prepareSplit();
+    uint64_t extraBytesToAdd{4};
+    writesTracker().addBytesWritten(extraBytesToAdd);
+    splitDriver()->cancelSplit();
+
+    ASSERT_EQ(writesTracker().getBytesWritten(), bytesInTracker + extraBytesToAdd);
+}
+
+TEST_F(ChunkSplitStateDriverTest,
+       PrepareSplitThenAddBytesThenCommitSplitLeavesNewBytesWrittenUnchanged) {
+    splitDriver()->prepareSplit();
+    uint64_t extraBytesToAdd{4};
+    writesTracker().addBytesWritten(extraBytesToAdd);
+    splitDriver()->commitSplit();
+
+    ASSERT_EQ(writesTracker().getBytesWritten(), extraBytesToAdd);
+}
+
+TEST_F(ChunkSplitStateDriverTest, ShouldSplitReturnsFalseWhenSplitHasBeenPrepared) {
+    splitDriver()->prepareSplit();
+
+    uint64_t maxChunkSize{0};
+    ASSERT_FALSE(writesTracker().shouldSplit(maxChunkSize));
+}
+
+TEST_F(ChunkSplitStateDriverTest, ShouldSplitReturnsFalseEvenAfterCommit) {
+    splitDriver()->prepareSplit();
+    splitDriver()->commitSplit();
+
+    uint64_t maxChunkSize{0};
+    ASSERT_FALSE(writesTracker().shouldSplit(maxChunkSize));
+}
+
+TEST_F(ChunkSplitStateDriverTest, ShouldSplitReturnsTrueAgainAfterCancel) {
+    splitDriver()->prepareSplit();
+    splitDriver()->cancelSplit();
+
+    uint64_t maxChunkSize{0};
+    ASSERT_TRUE(writesTracker().shouldSplit(maxChunkSize));
+}
+
+DEATH_TEST_F(ChunkSplitStateDriverTest,
+             CommitSplitWhenStartedAndNotPreparedErrors,
+             "Invariant failure") {
+    splitDriver()->commitSplit();
+}
+
+DEATH_TEST_F(ChunkSplitStateDriverTest, CancelSplitAfterCommitErrors, "Invariant failure") {
+    splitDriver()->commitSplit();
+    splitDriver()->cancelSplit();
+}
+
+}  // namespace mongo
diff --git a/src/mongo/s/chunk_writes_tracker.cpp b/src/mongo/s/chunk_writes_tracker.cpp
new file mode 100644
index 00000000000..35bb0bf3fe4
--- /dev/null
+++ b/src/mongo/s/chunk_writes_tracker.cpp
@@ -0,0 +1,68 @@
+/**
+ *    Copyright (C) 2018 MongoDB Inc.
+ *
+ *    This program is free software: you can redistribute it and/or  modify
+ *    it under the terms of the GNU Affero General Public License, version 3,
+ *    as published by the Free Software Foundation.
+ *
+ *    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
+ *    GNU Affero General Public License for more details.
+ *
+ *    You should have received a copy of the GNU Affero General Public License
+ *    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ *    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 GNU Affero General 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 <cstdint>
+
+#include "mongo/s/chunk_writes_tracker.h"
+#include "mongo/util/assert_util.h"
+
+namespace mongo {
+
+void ChunkWritesTracker::clearBytesWritten() {
+    _bytesWritten.store(0);
+}
+
+bool ChunkWritesTracker::shouldSplit(uint64_t maxChunkSize) {
+    if (_isLockedForSplitting) {
+        return false;
+    }
+
+    // Check if there are enough estimated bytes written to warrant a split
+    return getBytesWritten() > maxChunkSize / ChunkWritesTracker::kSplitTestFactor;
+}
+
+bool ChunkWritesTracker::acquireSplitLock() {
+    stdx::lock_guard<stdx::mutex> lk(_mtx);
+
+    if (!_isLockedForSplitting) {
+        _isLockedForSplitting = true;
+        return true;
+    }
+    return false;
+}
+
+void ChunkWritesTracker::releaseSplitLock() {
+    invariant(_isLockedForSplitting);
+    _isLockedForSplitting = false;
+}
+
+}  // namespace mongo
diff --git a/src/mongo/s/chunk_writes_tracker.h b/src/mongo/s/chunk_writes_tracker.h
new file mode 100644
index 00000000000..0511b3bc7fa
--- /dev/null
+++ b/src/mongo/s/chunk_writes_tracker.h
@@ -0,0 +1,99 @@
+/**
+ *    Copyright (C) 2018 MongoDB Inc.
+ *
+ *    This program is free software: you can redistribute it and/or  modify
+ *    it under the terms of the GNU Affero General Public License, version 3,
+ *    as published by the Free Software Foundation.
+ *
+ *    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
+ *    GNU Affero General Public License for more details.
+ *
+ *    You should have received a copy of the GNU Affero General Public License
+ *    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ *    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 GNU Affero General 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.
+ */
+
+#pragma once
+
+#include "mongo/base/disallow_copying.h"
+#include "mongo/platform/atomic_word.h"
+#include "mongo/stdx/mutex.h"
+
+namespace mongo {
+
+class ChunkWritesTracker {
+public:
+    /**
+     * A factor that determines when a chunk should be split. We should split once data *
+     * kSplitTestFactor > chunkSize (approximately).
+     */
+    static constexpr uint64_t kSplitTestFactor = 5;
+
+    /**
+     * Add more bytes written to the chunk.
+     */
+    void addBytesWritten(uint64_t bytesWritten) {
+        _bytesWritten.fetchAndAdd(bytesWritten);
+    }
+
+    /**
+     * Returns the total number of bytes that have been written to the chunk.
+     */
+    uint64_t getBytesWritten() {
+        return _bytesWritten.loadRelaxed();
+    }
+
+    /**
+     * Sets the number of bytes in the tracker to zero.
+     */
+    void clearBytesWritten();
+
+    /**
+     * Returns whether or not this chunk is ready to be split based on the
+     * maximum allowable size of a chunk.
+     */
+    bool shouldSplit(uint64_t maxChunkSize);
+
+    /**
+     * Locks the chunk for splitting, returning false if it is already locked.
+     * While it is locked, shouldSplit will always return false.
+     */
+    bool acquireSplitLock();
+
+    /**
+     * Releases the lock acquired for splitting.
+     */
+    void releaseSplitLock();
+
+private:
+    /**
+     * The number of bytes that have been written to this chunk. May be
+     * modified concurrently by several threads.
+     */
+    AtomicUInt64 _bytesWritten{0};
+
+    /**
+     * Protects _splitState when starting a split.
+     */
+    stdx::mutex _mtx;
+
+    /**
+     * Whether or not a current split is in progress for this chunk.
+     */
+    bool _isLockedForSplitting{false};
+};
+
+}  // namespace mongo
diff --git a/src/mongo/s/chunk_writes_tracker_test.cpp b/src/mongo/s/chunk_writes_tracker_test.cpp
new file mode 100644
index 00000000000..098397ca11c
--- /dev/null
+++ b/src/mongo/s/chunk_writes_tracker_test.cpp
@@ -0,0 +1,143 @@
+/**
+ * Copyright (C) 2018 MongoDB Inc.
+ *
+ * This program is free software: you can redistribute it and/or  modify
+ * it under the terms of the GNU Affero General Public License, version 3,
+ * as published by the Free Software Foundation.
+ *
+ * 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
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ *
+ * 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 GNU Affero General 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.
+ */
+
+#include "mongo/platform/basic.h"
+
+#include "mongo/s/chunk_writes_tracker.h"
+
+#include "mongo/unittest/death_test.h"
+#include "mongo/unittest/unittest.h"
+
+
+namespace mongo {
+
+TEST(ChunkWritesTrackerTest, BytesWrittenStartsAtZero) {
+    ChunkWritesTracker wt;
+    ASSERT_EQ(wt.getBytesWritten(), 0ull);
+}
+
+TEST(ChunkWritesTrackerTest, AddBytesWrittenCorrectlyAddsBytes) {
+    ChunkWritesTracker wt;
+    uint64_t bytesToAdd{4};
+    wt.addBytesWritten(bytesToAdd);
+    ASSERT_EQ(wt.getBytesWritten(), bytesToAdd);
+}
+
+TEST(ChunkWritesTrackerTest, ClearBytesWrittenSetsBytesToZero) {
+    ChunkWritesTracker wt;
+    wt.addBytesWritten(4ull);
+    wt.clearBytesWritten();
+    ASSERT_EQ(wt.getBytesWritten(), 0ull);
+}
+
+TEST(ChunkWritesTrackerTest, ShouldSplitReturnsTrueWithBytesWrittenAndMaxChunkSizeZero) {
+    ChunkWritesTracker wt;
+    wt.addBytesWritten(4ull);
+    uint64_t maxChunkSize{0};
+    ASSERT_TRUE(wt.shouldSplit(maxChunkSize));
+}
+
+TEST(ChunkWritesTrackerTest, ShouldSplitReturnsFalseWithNoBytesWrittenAndMaxChunkSizeZero) {
+    ChunkWritesTracker wt;
+    uint64_t maxChunkSize{0};
+    ASSERT_FALSE(wt.shouldSplit(maxChunkSize));
+}
+
+TEST(ChunkWritesTrackerTest, ShouldSplitReturnsTrueWithBytesWrittenGreaterThanMaxChunkSize) {
+    ChunkWritesTracker wt;
+    wt.addBytesWritten(4ull);
+    uint64_t maxChunkSize{3};
+    ASSERT_TRUE(wt.shouldSplit(maxChunkSize));
+}
+
+TEST(ChunkWritesTrackerTest, ShouldSplitReturnsFalseWithBytesWrittenLessThanThreshold) {
+    ChunkWritesTracker wt;
+    uint64_t maxChunkSize{10};
+    auto expectedCutoff = maxChunkSize / ChunkWritesTracker::kSplitTestFactor;
+    wt.addBytesWritten(expectedCutoff - 1);
+    ASSERT_FALSE(wt.shouldSplit(maxChunkSize));
+    wt.addBytesWritten(1ul);
+    ASSERT_FALSE(wt.shouldSplit(maxChunkSize));
+    wt.addBytesWritten(1ul);
+    ASSERT_TRUE(wt.shouldSplit(maxChunkSize));
+}
+
+TEST(ChunkWritesTrackerTest, ShouldSplitReturnsFalseWithBytesWrittenEqualToThreshold) {
+    ChunkWritesTracker wt;
+    uint64_t maxChunkSize{10};
+    auto expectedCutoff = maxChunkSize / ChunkWritesTracker::kSplitTestFactor;
+    wt.addBytesWritten(expectedCutoff);
+    ASSERT_FALSE(wt.shouldSplit(maxChunkSize));
+}
+
+TEST(ChunkWritesTrackerTest, ShouldSplitReturnsFalseWithBytesWrittenGreaterThanThreshold) {
+    ChunkWritesTracker wt;
+    uint64_t maxChunkSize{10};
+    auto expectedCutoff = maxChunkSize / ChunkWritesTracker::kSplitTestFactor;
+    wt.addBytesWritten(expectedCutoff + 1);
+    ASSERT_TRUE(wt.shouldSplit(maxChunkSize));
+}
+
+TEST(ChunkWritesTrackerTest, ShouldSplitReturnsFalseWhenSplitLockAcquired) {
+    ChunkWritesTracker wt;
+    wt.addBytesWritten(4ull);
+    wt.acquireSplitLock();
+    uint64_t maxChunkSize{0};
+    ASSERT_FALSE(wt.shouldSplit(maxChunkSize));
+}
+
+TEST(ChunkWritesTrackerTest, ShouldSplitReturnsTrueAfterSplitLockReleased) {
+    ChunkWritesTracker wt;
+    wt.addBytesWritten(4ull);
+    wt.acquireSplitLock();
+    wt.releaseSplitLock();
+    uint64_t maxChunkSize{0};
+    ASSERT_TRUE(wt.shouldSplit(maxChunkSize));
+}
+
+TEST(ChunkWritesTrackerTest, AcquireSplitLockReturnsFalseAfterReturningTrue) {
+    ChunkWritesTracker wt;
+    wt.addBytesWritten(4ull);
+    ASSERT_TRUE(wt.acquireSplitLock());
+    ASSERT_FALSE(wt.acquireSplitLock());
+}
+
+TEST(ChunkWritesTrackerTest, AcquireSplitLockThenReleaseThenReacquireReturnsTrue) {
+    ChunkWritesTracker wt;
+    wt.addBytesWritten(4ull);
+    wt.acquireSplitLock();
+    wt.releaseSplitLock();
+    ASSERT_TRUE(wt.acquireSplitLock());
+}
+
+DEATH_TEST(ChunkWritesTrackerTest, ReleaseSplitLockWithoutAcquiringErrors, "Invariant failure") {
+    ChunkWritesTracker wt;
+    wt.releaseSplitLock();
+}
+
+}  // namespace mongo
diff --git a/src/mongo/s/write_ops/cluster_write.cpp b/src/mongo/s/write_ops/cluster_write.cpp
index 18850598cf0..19311ea39ab 100644
--- a/src/mongo/s/write_ops/cluster_write.cpp
+++ b/src/mongo/s/write_ops/cluster_write.cpp
@@ -237,7 +237,7 @@ void ClusterWriter::write(OperationContext* opCtx,
 void updateChunkWriteStatsAndSplitIfNeeded(OperationContext* opCtx,
                                            ChunkManager* manager,
                                            Chunk chunk,
-                                           long dataWritten) {
+                                           long chunkBytesWritten) {
     // Disable lastError tracking so that any errors, which occur during auto-split do not get
     // bubbled up on the client connection doing a write
     LastError::Disabled disableLastError(&LastError::get(opCtx->getClient()));
@@ -249,11 +249,11 @@ void updateChunkWriteStatsAndSplitIfNeeded(OperationContext* opCtx,
     const bool maxIsInf =
         (0 == manager->getShardKeyPattern().getKeyPattern().globalMax().woCompare(chunk.getMax()));
 
-    const uint64_t chunkBytesWritten = chunk.addBytesWritten(dataWritten);
+    chunk.addBytesWritten(chunkBytesWritten);
 
     const uint64_t desiredChunkSize = balancerConfig->getMaxChunkSizeBytes();
 
-    if (!chunk.shouldSplit(desiredChunkSize, minIsInf, maxIsInf)) {
+    if (!chunk.shouldSplit(desiredChunkSize)) {
         return;
     }
 
@@ -287,7 +287,8 @@ void updateChunkWriteStatsAndSplitIfNeeded(OperationContext* opCtx,
                 // The current desired chunk size will split the chunk into lots of small chunk and
                 // at the worst case this can result into thousands of chunks. So check and see if a
                 // bigger value can be used.
-                return std::min(chunkBytesWritten, balancerConfig->getMaxChunkSizeBytes());
+                return std::min((uint64_t)chunkBytesWritten,
+                                balancerConfig->getMaxChunkSizeBytes());
             } else {
                 return desiredChunkSize;
             }