SERVER-19583: Implement full-time diagnostic data capture compressor and decompressor

16 files changed, 2322 insertions, 0 deletions

diff --git a/src/mongo/db/ftdc/SConscript b/src/mongo/db/ftdc/SConscript
new file mode 100644
index 00000000000..b050d88f0bb
--- /dev/null
+++ b/src/mongo/db/ftdc/SConscript
@@ -0,0 +1,36 @@
+# -*- mode: python -*-
+Import("env")
+
+ftdcEnv = env.Clone()
+ftdcEnv.InjectThirdPartyIncludePaths(libraries=['zlib'])
+
+ftdcEnv.Library(
+    target='ftdc',
+    source=[
+        'block_compressor.cpp',
+        'compressor.cpp',
+        'decompressor.cpp',
+        'util.cpp',
+        'varint.cpp'
+    ],
+    LIBDEPS=[
+        '$BUILD_DIR/mongo/base',
+        '$BUILD_DIR/mongo/db/service_context',
+        '$BUILD_DIR/third_party/s2/s2', # For VarInt
+        '$BUILD_DIR/third_party/shim_zlib',
+    ],
+)
+
+env.CppUnitTest(
+    target='ftdc_test',
+    source=[
+        'compressor_test.cpp',
+        'ftdc_test.cpp',
+        'util_test.cpp',
+        'varint_test.cpp',
+    ],
+    LIBDEPS=[
+        '$BUILD_DIR/mongo/bson/util/bson_extract',
+        'ftdc',
+    ])
+
diff --git a/src/mongo/db/ftdc/block_compressor.cpp b/src/mongo/db/ftdc/block_compressor.cpp
new file mode 100644
index 00000000000..1077426fc4b
--- /dev/null
+++ b/src/mongo/db/ftdc/block_compressor.cpp
@@ -0,0 +1,119 @@
+/**
+ * Copyright (C) 2015 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/db/ftdc/block_compressor.h"
+
+#include <zlib.h>
+
+#include "mongo/util/mongoutils/str.h"
+
+namespace mongo {
+
+StatusWith<ConstDataRange> BlockCompressor::compress(ConstDataRange source) {
+    z_stream stream;
+    int level = Z_DEFAULT_COMPRESSION;
+
+    stream.next_in = reinterpret_cast<unsigned char*>(const_cast<char*>(source.data()));
+    stream.avail_in = source.length();
+
+    // In compress.c in the zlib source, they recommend that the compression buffer be
+    // at least 0.1% larger + 12 bytes then the source length.
+    // We make the buffer 1% larger then the source length buffer to be on the safe side. If we are
+    // too small, deflate returns an error.
+    _buffer.resize(source.length() * 1.01 + 12);
+
+    stream.next_out = _buffer.data();
+    stream.avail_out = _buffer.size();
+
+    stream.zalloc = nullptr;
+    stream.zfree = nullptr;
+    stream.opaque = nullptr;
+
+    int err = deflateInit(&stream, level);
+    if (err != Z_OK) {
+        return {ErrorCodes::ZLibError, str::stream() << "deflateInit failed with " << err};
+    }
+
+    err = deflate(&stream, Z_FINISH);
+    if (err != Z_STREAM_END) {
+        (void)deflateEnd(&stream);
+
+        if (err != Z_OK) {
+            return {ErrorCodes::ZLibError, str::stream() << "deflate failed with " << err};
+        }
+    }
+
+    err = deflateEnd(&stream);
+    if (err != Z_OK) {
+        return {ErrorCodes::ZLibError, str::stream() << "deflateEnd failed with " << err};
+    }
+
+    return ConstDataRange(reinterpret_cast<char*>(_buffer.data()), stream.total_out);
+}
+
+StatusWith<ConstDataRange> BlockCompressor::uncompress(ConstDataRange source,
+                                                       size_t uncompressedLength) {
+    z_stream stream;
+
+    stream.next_in = reinterpret_cast<unsigned char*>(const_cast<char*>(source.data()));
+    stream.avail_in = source.length();
+
+    _buffer.resize(uncompressedLength);
+
+    stream.next_out = _buffer.data();
+    stream.avail_out = _buffer.size();
+
+    stream.zalloc = nullptr;
+    stream.zfree = nullptr;
+    stream.opaque = nullptr;
+
+    int err = inflateInit(&stream);
+    if (err != Z_OK) {
+        return {ErrorCodes::ZLibError, str::stream() << "inflateInit failed with " << err};
+    }
+
+    err = inflate(&stream, Z_FINISH);
+    if (err != Z_STREAM_END) {
+        (void)inflateEnd(&stream);
+
+        if (err != Z_OK) {
+            return {ErrorCodes::ZLibError, str::stream() << "inflate failed with " << err};
+        }
+    }
+
+    err = inflateEnd(&stream);
+    if (err != Z_OK) {
+        return {ErrorCodes::ZLibError, str::stream() << "inflateEnd failed with " << err};
+    }
+
+    return ConstDataRange(reinterpret_cast<char*>(_buffer.data()), stream.total_out);
+}
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/block_compressor.h b/src/mongo/db/ftdc/block_compressor.h
new file mode 100644
index 00000000000..23ac8c811c0
--- /dev/null
+++ b/src/mongo/db/ftdc/block_compressor.h
@@ -0,0 +1,72 @@
+/**
+ * Copyright (C) 2015 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 <cstdint>
+#include <vector>
+
+#include "mongo/base/data_range.h"
+#include "mongo/base/disallow_copying.h"
+#include "mongo/base/status_with.h"
+
+namespace mongo {
+
+/**
+ * Compesses and uncompresses a block of buffer using zlib.
+ */
+class BlockCompressor {
+    MONGO_DISALLOW_COPYING(BlockCompressor);
+
+public:
+    BlockCompressor() = default;
+
+    /**
+     * Compress a buffer of data.
+     *
+     * Returns a pointer to a buffer that BlockCompressor owns.
+     * The returned buffer is valid until the next call to compress or uncompress.
+     */
+    StatusWith<ConstDataRange> compress(ConstDataRange source);
+
+    /**
+     * Uncompress a buffer of data.
+     *
+     * maxUncompressedLength is the upper bound on the size of the uncompressed data
+     * so that an internal buffer can be allocated to fit it.
+     *
+     * Returns a pointer to a buffer that BlockCompressor owns.
+     * The returned buffer is valid until the next call to compress or uncompress.
+     */
+    StatusWith<ConstDataRange> uncompress(ConstDataRange source, size_t maxUncompressedLength);
+
+private:
+    std::vector<std::uint8_t> _buffer;
+};
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/compressor.cpp b/src/mongo/db/ftdc/compressor.cpp
new file mode 100644
index 00000000000..1467075ed1c
--- /dev/null
+++ b/src/mongo/db/ftdc/compressor.cpp
@@ -0,0 +1,218 @@
+/**
+ * Copyright (C) 2015 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/db/ftdc/compressor.h"
+
+#include "mongo/base/data_builder.h"
+#include "mongo/db/ftdc/config.h"
+#include "mongo/db/ftdc/util.h"
+#include "mongo/db/ftdc/varint.h"
+#include "mongo/db/jsobj.h"
+#include "mongo/util/assert_util.h"
+
+namespace mongo {
+
+using std::swap;
+
+StatusWith<boost::optional<std::tuple<ConstDataRange, FTDCCompressor::CompressorState>>>
+FTDCCompressor::addSample(const BSONObj& sample) {
+    if (_referenceDoc.isEmpty()) {
+        FTDCBSONUtil::extractMetricsFromDocument(sample, sample, &_metrics);
+        _reset(sample);
+        return {boost::none_t()};
+    }
+
+    _metrics.resize(0);
+
+    auto swMatches = FTDCBSONUtil::extractMetricsFromDocument(_referenceDoc, sample, &_metrics);
+
+    if (!swMatches.isOK()) {
+        return swMatches.getStatus();
+    }
+
+    dassert((swMatches.getValue() == false || _metricsCount == _metrics.size()) &&
+            _metrics.size() < std::numeric_limits<std::uint32_t>::max());
+
+    // We need to flush the current set of samples since the BSON schema has changed.
+    if (!swMatches.getValue()) {
+        auto swCompressedSamples = getCompressedSamples();
+
+        if (!swCompressedSamples.isOK()) {
+            return swCompressedSamples.getStatus();
+        }
+
+        // Set the new sample as the current reference document as we have to start all over
+        _reset(sample);
+        return {boost::optional<std::tuple<ConstDataRange, FTDCCompressor::CompressorState>>(
+            std::tuple<ConstDataRange, FTDCCompressor::CompressorState>(
+                swCompressedSamples.getValue(), CompressorState::kSchemaChanged))};
+    }
+
+
+    // Add another sample
+    for (std::size_t i = 0; i < _metrics.size(); ++i) {
+        // NOTE: This touches a lot of cache lines so that compression code can be more effcient.
+        _deltas[getArrayOffset(_maxSamples, _sampleCount, i)] = _metrics[i] - _prevmetrics[i];
+    }
+
+    ++_sampleCount;
+
+    _prevmetrics.clear();
+    swap(_prevmetrics, _metrics);
+
+    // If the count is full, flush
+    if (_sampleCount == _maxSamples) {
+        auto swCompressedSamples = getCompressedSamples();
+
+        if (!swCompressedSamples.isOK()) {
+            return swCompressedSamples.getStatus();
+        }
+
+        // Setup so that we treat the next sample as the reference sample
+        _referenceDoc = BSONObj();
+
+        return {boost::optional<std::tuple<ConstDataRange, FTDCCompressor::CompressorState>>(
+            std::tuple<ConstDataRange, FTDCCompressor::CompressorState>(
+                swCompressedSamples.getValue(), CompressorState::kCompressorFull))};
+    }
+
+    // The buffer is not full, inform the caller
+    return {boost::none_t()};
+}
+
+StatusWith<ConstDataRange> FTDCCompressor::getCompressedSamples() {
+    _chunkBuffer.setlen(0);
+
+    // Append reference document - BSON Object
+    _chunkBuffer.appendBuf(_referenceDoc.objdata(), _referenceDoc.objsize());
+
+    // Append count of metrics - uint32 little endian
+    _chunkBuffer.appendNum(static_cast<std::uint32_t>(_metricsCount));
+
+    // Append count of samples - uint32 little endian
+    _chunkBuffer.appendNum(static_cast<std::uint32_t>(_sampleCount));
+
+    if (_metricsCount == 0 || _sampleCount == 0) {
+        return ConstDataRange(_chunkBuffer.buf(), static_cast<size_t>(_chunkBuffer.len()));
+    }
+
+    // On average, we do not need all 10 bytes for every sample, worst case, we grow the buffer
+    DataBuilder db(_metricsCount * _sampleCount * FTDCVarInt::kMaxSizeBytes64 / 2);
+
+    std::uint32_t zeroesCount = 0;
+
+    // For each set of samples for a particular metric,
+    // we think of it is simple array of 64-bit integers we try to compress into a byte array.
+    // This is done in three steps for each metric
+    // 1. Delta Compression
+    //   - i.e., we store the difference between pairs of samples, not their absolute values
+    //   - this is done in addSamples
+    // 2. Run Length Encoding of zeros
+    //   - We find consecutive sets of zeros and represent them as a tuple of (0, count - 1).
+    //   - Each memeber is stored as VarInt packed integer
+    // 3. Finally, for non-zero members, we store these as VarInt packed
+    //
+    // These byte arrays are added to a buffer which is then concatenated with other chunks and
+    // compressed with ZLIB.
+    for (std::uint32_t i = 0; i < _metricsCount; i++) {
+        for (std::uint32_t j = 0; j < _sampleCount; j++) {
+            std::uint64_t delta = _deltas[getArrayOffset(_maxSamples, j, i)];
+
+            if (delta == 0) {
+                ++zeroesCount;
+                continue;
+            }
+
+            // If we have a non-zero sample, then write out all the accumulated zero samples.
+            if (zeroesCount > 0) {
+                auto s1 = db.writeAndAdvance(FTDCVarInt(0));
+                if (!s1.isOK()) {
+                    return s1;
+                }
+
+                auto s2 = db.writeAndAdvance(FTDCVarInt(zeroesCount - 1));
+                if (!s2.isOK()) {
+                    return s2;
+                }
+
+                zeroesCount = 0;
+            }
+
+            auto s3 = db.writeAndAdvance(FTDCVarInt(delta));
+            if (!s3.isOK()) {
+                return s3;
+            }
+        }
+
+        // If we are on the last metric, and the previous loop ended in a zero, write out the RLE
+        // pair of zero information.
+        if ((i == (_metricsCount - 1)) && zeroesCount) {
+            auto s1 = db.writeAndAdvance(FTDCVarInt(0));
+            if (!s1.isOK()) {
+                return s1;
+            }
+
+            auto s2 = db.writeAndAdvance(FTDCVarInt(zeroesCount - 1));
+            if (!s2.isOK()) {
+                return s2;
+            }
+        }
+    }
+
+    auto swDest = _compressor.compress(db.getCursor());
+
+    // The only way for compression to fail is if the buffer size calculations are wrong
+    if (!swDest.isOK()) {
+        return swDest.getStatus();
+    }
+
+    _chunkBuffer.appendBuf(swDest.getValue().data(), swDest.getValue().length());
+
+    return ConstDataRange(_chunkBuffer.buf(), static_cast<size_t>(_chunkBuffer.len()));
+}
+
+void FTDCCompressor::reset() {
+    _metrics.clear();
+    _reset(BSONObj());
+}
+
+void FTDCCompressor::_reset(const BSONObj& referenceDoc) {
+    _referenceDoc = referenceDoc;
+
+    _metricsCount = _metrics.size();
+    _sampleCount = 0;
+    _prevmetrics.clear();
+    swap(_prevmetrics, _metrics);
+
+    _maxSamples = _config->maxSamplesPerArchiveMetricChunk;
+    _deltas.resize(_metricsCount * _maxSamples);
+}
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/compressor.h b/src/mongo/db/ftdc/compressor.h
new file mode 100644
index 00000000000..e1a994ccf2d
--- /dev/null
+++ b/src/mongo/db/ftdc/compressor.h
@@ -0,0 +1,177 @@
+/**
+ * Copyright (C) 2015 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 <cstddef>
+#include <cstdint>
+#include <tuple>
+#include <vector>
+
+#include "mongo/base/disallow_copying.h"
+#include "mongo/base/status_with.h"
+#include "mongo/bson/util/builder.h"
+#include "mongo/db/ftdc/block_compressor.h"
+#include "mongo/db/ftdc/config.h"
+#include "mongo/db/jsobj.h"
+
+namespace mongo {
+
+/**
+ * FTDCCompressor is responsible for taking a set of BSON documents containing metrics, and
+ * compressing them into a highly compressed buffer. Metrics are defined as BSON number or number
+ * like type (like dates, and timestamps).
+ *
+ * Compression Method
+ * 1. For each document after the first, it computes the delta between it and the preceding document
+ *    for the number fields
+ * 2. It stores the deltas into an array of std::int64_t.
+ * 3. It compressed each std::int64_t using VarInt integer compression. See varint.h.
+ * 4. Encodes zeros in Run Length Encoded pairs of <Count, Zero>
+ * 5. ZLIB compresses the final processed array
+ *
+ * NOTE: This compression ignores non-number data, and assumes the non-number data is constant
+ * across all documents in the series of documents.
+ */
+class FTDCCompressor {
+    MONGO_DISALLOW_COPYING(FTDCCompressor);
+
+public:
+    /**
+     * The FTDCCompressor returns one of these values when a sample is added to indicate whether the
+     * caller should flush the buffer to disk or not.
+     */
+    enum class CompressorState {
+        /**
+         * Needs to flush because the schemas has changed. Caller needs to flush.
+         */
+        kSchemaChanged,
+
+        /**
+         * Quota on the number of samples in a metric chunk has been reached. Caller needs to flush.
+         */
+        kCompressorFull,
+    };
+
+    explicit FTDCCompressor(const FTDCConfig* config) : _config(config) {}
+
+    /**
+     * Add a bson document containing metrics into the compressor.
+     *
+     * Returns flag indicating whether the caller should flush the compressor buffer to disk.
+     *  1. kCompressorFull if the compressor is considered full.
+     *  2. kSchemaChanged if there was a schema change, and buffer should be flushed.
+     *  3. kHasSpace if it has room for more metrics in the current buffer.
+     */
+    StatusWith<boost::optional<std::tuple<ConstDataRange, CompressorState>>> addSample(
+        const BSONObj& sample);
+
+    /**
+     * Returns the number of enqueued samples.
+     *
+     * The a buffer will decompress to (1 + getCountCount). The extra 1 comes
+     * from the reference document.
+     */
+    std::size_t getSampleCount() const {
+        return _sampleCount;
+    }
+
+    /**
+     * Has a document been added?
+     *
+     * If addSample has been called, then we have at least the reference document, but not
+     * necessarily any additional metric samples. When the buffer is filled to capacity,
+     * the reference document is reset.
+     */
+    bool hasDataToFlush() const {
+        return !_referenceDoc.isEmpty() || _chunkBuffer.len() > 0;
+    }
+
+    /**
+     * Gets buffer of compressed data contained in the FTDCCompressor.
+     *
+     * The returned buffer is valid until next call to addSample() or getCompressedSamples() with
+     * CompressBuffer::kGenerateNewCompressedBuffer.
+     */
+    StatusWith<ConstDataRange> getCompressedSamples();
+
+    /**
+     * Reset the state of the compressor.
+     *
+     * Callers can use this to reset the compressor to a clean state instead of recreating it.
+     */
+    void reset();
+
+    /**
+     * Compute the offset into an array for given (sample, metric) pair
+     */
+    static size_t getArrayOffset(std::uint32_t sampleCount,
+                                 std::uint32_t sample,
+                                 std::uint32_t metric) {
+        return metric * sampleCount + sample;
+    }
+
+private:
+    /**
+     * Reset the state
+     */
+    void _reset(const BSONObj& referenceDoc);
+
+private:
+    // Block Compressor
+    BlockCompressor _compressor;
+
+    // Config
+    const FTDCConfig* const _config;
+
+    // Reference schema document
+    BSONObj _referenceDoc;
+
+    // Number of Metrics for the reference document
+    std::uint32_t _metricsCount{0};
+
+    // Number of samples recorded
+    std::uint32_t _sampleCount{0};
+
+    // Max samples for the current chunk
+    std::size_t _maxSamples{0};
+
+    // Array of deltas - M x S
+    // _deltas[Metrics][Samples]
+    std::vector<std::uint64_t> _deltas;
+
+    // Buffer for metric chunk = header + zlib compressed array
+    BufBuilder _chunkBuffer;
+
+    // Buffer to hold metrics
+    std::vector<std::uint64_t> _metrics;
+    std::vector<std::uint64_t> _prevmetrics;
+};
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/compressor_test.cpp b/src/mongo/db/ftdc/compressor_test.cpp
new file mode 100644
index 00000000000..7b423679855
--- /dev/null
+++ b/src/mongo/db/ftdc/compressor_test.cpp
@@ -0,0 +1,429 @@
+/**
+ * Copyright (C) 2015 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 <limits>
+#include <random>
+
+#include "mongo/base/status_with.h"
+#include "mongo/bson/bsonmisc.h"
+#include "mongo/bson/bsonobjbuilder.h"
+#include "mongo/db/ftdc/compressor.h"
+#include "mongo/db/ftdc/config.h"
+#include "mongo/db/ftdc/decompressor.h"
+#include "mongo/db/ftdc/ftdc_test.h"
+#include "mongo/db/jsobj.h"
+#include "mongo/unittest/unittest.h"
+#include "mongo/util/assert_util.h"
+
+namespace mongo {
+
+#define ASSERT_HAS_SPACE(st) \
+    ASSERT_TRUE(st.isOK());  \
+    ASSERT_FALSE(st.getValue().is_initialized());
+
+#define ASSERT_SCHEMA_CHANGED(st)                                 \
+    ASSERT_TRUE(st.isOK());                                       \
+    ASSERT_TRUE(std::get<1>(st.getValue().get()) ==               \
+                FTDCCompressor::CompressorState::kSchemaChanged); \
+    ASSERT_TRUE(st.getValue().is_initialized());
+
+#define ASSERT_FULL(st)                                            \
+    ASSERT_TRUE(st.isOK());                                        \
+    ASSERT_TRUE(std::get<1>(st.getValue().get()) ==                \
+                FTDCCompressor::CompressorState::kCompressorFull); \
+    ASSERT_TRUE(st.getValue().is_initialized());
+
+// Sanity check
+TEST(FTDCCompressor, TestBasic) {
+    FTDCConfig config;
+    FTDCCompressor c(&config);
+
+    auto st = c.addSample(BSON("name"
+                               << "joe"
+                               << "key1" << 33 << "key2" << 42));
+    ASSERT_HAS_SPACE(st);
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34 << "key2" << 45));
+    ASSERT_HAS_SPACE(st);
+
+
+    StatusWith<ConstDataRange> swBuf = c.getCompressedSamples();
+
+    ASSERT_TRUE(swBuf.isOK());
+    ASSERT_TRUE(swBuf.getValue().length() > 0);
+    ASSERT_TRUE(swBuf.getValue().data() != nullptr);
+}
+
+// Test strings only
+TEST(FTDCCompressor, TestStrings) {
+    FTDCConfig config;
+    FTDCCompressor c(&config);
+
+    auto st = c.addSample(BSON("name"
+                               << "joe"
+                               << "key1"
+                               << "value1"
+                               << "key2"
+                               << "value2"));
+    ASSERT_HAS_SPACE(st);
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1"
+                          << "value3"
+                          << "key2"
+                          << "value6"));
+    ASSERT_HAS_SPACE(st);
+
+    StatusWith<ConstDataRange> swBuf = c.getCompressedSamples();
+
+    ASSERT_TRUE(swBuf.isOK());
+    ASSERT_TRUE(swBuf.getValue().length() > 0);
+    ASSERT_TRUE(swBuf.getValue().data() != nullptr);
+}
+
+/**
+ * Test class that records a series of samples and ensures that compress + decompress round trips
+ * them correctly.
+ */
+class TestTie {
+public:
+    TestTie() : _compressor(&_config) {}
+
+    ~TestTie() {
+        validate(boost::none_t());
+    }
+
+    StatusWith<boost::optional<std::tuple<ConstDataRange, FTDCCompressor::CompressorState>>>
+    addSample(const BSONObj& sample) {
+        auto st = _compressor.addSample(sample);
+
+        if (!st.getValue().is_initialized()) {
+            _docs.emplace_back(sample);
+        } else if (std::get<1>(st.getValue().get()) ==
+                   FTDCCompressor::CompressorState::kSchemaChanged) {
+            validate(std::get<0>(st.getValue().get()));
+            _docs.clear();
+            _docs.emplace_back(sample);
+        } else if (std::get<1>(st.getValue().get()) ==
+                   FTDCCompressor::CompressorState::kCompressorFull) {
+            _docs.emplace_back(sample);
+            validate(std::get<0>(st.getValue().get()));
+            _docs.clear();
+        } else {
+            MONGO_UNREACHABLE;
+        }
+
+        return st;
+    }
+
+    void validate(boost::optional<ConstDataRange> cdr) {
+        std::vector<BSONObj> list;
+        if (cdr.is_initialized()) {
+            auto sw = _decompressor.uncompress(cdr.get());
+            ASSERT_TRUE(sw.isOK());
+            list = sw.getValue();
+        } else {
+            auto swBuf = _compressor.getCompressedSamples();
+            ASSERT_TRUE(swBuf.isOK());
+            auto sw = _decompressor.uncompress(swBuf.getValue());
+            ASSERT_TRUE(sw.isOK());
+
+            list = sw.getValue();
+        }
+
+        ValidateDocumentList(list, _docs);
+    }
+
+private:
+    std::vector<BSONObj> _docs;
+    FTDCConfig _config;
+    FTDCCompressor _compressor;
+    FTDCDecompressor _decompressor;
+};
+
+// Test various schema changes
+TEST(FTDCCompressor, TestSchemaChanges) {
+    TestTie c;
+
+    auto st = c.addSample(BSON("name"
+                               << "joe"
+                               << "key1" << 33 << "key2" << 42));
+    ASSERT_HAS_SPACE(st);
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34 << "key2" << 45));
+    ASSERT_HAS_SPACE(st);
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34 << "key2" << 45));
+    ASSERT_HAS_SPACE(st);
+
+    // Add Field
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34 << "key2" << 45 << "key3" << 47));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34 << "key2" << 45 << "key3" << 47));
+    ASSERT_HAS_SPACE(st);
+
+    // Rename field
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34 << "key5" << 45 << "key3" << 47));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Change type
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34 << "key5"
+                          << "45"
+                          << "key3" << 47));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Add Field
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34 << "key2" << 45 << "key3" << 47 << "key7" << 34 << "key9"
+                          << 45 << "key13" << 47));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Remove Field
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key7" << 34 << "key9" << 45 << "key13" << 47));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key7" << 34 << "key9" << 45 << "key13" << 47));
+    ASSERT_HAS_SPACE(st);
+
+    // Start new batch
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key7" << 5));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Change field to object
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key7" << BSON(  // nested object
+                                           "a" << 1)));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Change field from object to number
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key7" << 7));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Change field from number to array
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key7" << BSON_ARRAY(13 << 17)));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Change field from array to number
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key7" << 19));
+    ASSERT_SCHEMA_CHANGED(st);
+
+
+    // New Schema
+    st = c.addSample(BSON("_id" << 1));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Change field to oid
+    st = c.addSample(BSON(GENOID));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Change field from oid to object
+    st = c.addSample(BSON("_id" << BSON("sub1" << 1)));
+    ASSERT_SCHEMA_CHANGED(st);
+
+    // Change field from object to oid
+    st = c.addSample(BSON(GENOID));
+    ASSERT_SCHEMA_CHANGED(st);
+}
+
+// Ensure changing between the various number formats is considered compatible
+TEST(FTDCCompressor, TestNumbersCompat) {
+    TestTie c;
+
+    auto st = c.addSample(BSON("name"
+                               << "joe"
+                               << "key1" << 33 << "key2" << 42LL));
+    ASSERT_HAS_SPACE(st);
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34LL << "key2" << 45.0f));
+    ASSERT_HAS_SPACE(st);
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << static_cast<char>(32) << "key2" << 45.0F));
+    ASSERT_HAS_SPACE(st);
+}
+
+// Test timestamp
+TEST(FTDCCompressor, TestTimeStamp) {
+    TestTie c;
+
+    BSONObjBuilder builder1;
+    BSONObj o = builder1.append("ts", Timestamp(0x55667788LL, 0x11223344LL)).obj();
+    auto st = c.addSample(o);
+    ASSERT_HAS_SPACE(st);
+}
+
+// Test all types
+TEST(FTDCCompressor, Types) {
+    TestTie c;
+
+    auto st = c.addSample(BSON("name"
+                               << "joe"
+                               << "key1" << 33 << "key2" << 42LL));
+    ASSERT_HAS_SPACE(st);
+
+    const char bytes[] = {0x1, 0x2, 0x3};
+    BSONObj o = BSON("created" << DATENOW                       // date_t
+                               << "null" << BSONNULL            // { a : null }
+                               << "undefined" << BSONUndefined  // { a : undefined }
+                               << "obj" << BSON(                // nested object
+                                               "a"
+                                               << "abc"
+                                               << "b" << 123LL) << "foo"
+                               << BSON_ARRAY("bar"
+                                             << "baz"
+                                             << "qux")               // array of strings
+                               << "foo2" << BSON_ARRAY(5 << 6 << 7)  // array of ints
+                               << "bindata" << BSONBinData(&bytes[0], 3, bdtCustom)  // bindata
+                               << "oid" << OID("010203040506070809101112")           // oid
+                               << "bool" << true                                     // bool
+                               << "regex" << BSONRegEx("mongodb")                    // regex
+                               << "ref" << BSONDBRef("c", OID("010203040506070809101112"))  // ref
+                               << "code" << BSONCode("func f() { return 1; }")              // code
+                               << "codewscope" << BSONCodeWScope("func f() { return 1; }",
+                                                                 BSON("c" << true))  // codew
+                               << "minkey" << MINKEY                                 // minkey
+                               << "maxkey" << MAXKEY                                 // maxkey
+                     );
+
+    st = c.addSample(o);
+    ASSERT_SCHEMA_CHANGED(st);
+
+    st = c.addSample(o);
+    ASSERT_HAS_SPACE(st);
+
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << 34LL << "key2" << 45.0f));
+    ASSERT_SCHEMA_CHANGED(st);
+    st = c.addSample(BSON("name"
+                          << "joe"
+                          << "key1" << static_cast<char>(32) << "key2" << 45.0F));
+    ASSERT_HAS_SPACE(st);
+}
+
+// Test a full buffer
+TEST(FTDCCompressor, TestFull) {
+    // Test a large numbers of zeros, and incremental numbers in a full buffer
+    for (int j = 0; j < 2; j++) {
+        TestTie c;
+
+        auto st = c.addSample(BSON("name"
+                                   << "joe"
+                                   << "key1" << 33 << "key2" << 42));
+        ASSERT_HAS_SPACE(st);
+
+        for (size_t i = 0; i <= FTDCConfig::kMaxSamplesPerArchiveMetricChunkDefault - 2; i++) {
+            st = c.addSample(BSON("name"
+                                  << "joe"
+                                  << "key1" << static_cast<long long int>(i * j) << "key2" << 45));
+            ASSERT_HAS_SPACE(st);
+        }
+
+        st = c.addSample(BSON("name"
+                              << "joe"
+                              << "key1" << 34 << "key2" << 45));
+        ASSERT_FULL(st);
+
+        // Add Value
+        st = c.addSample(BSON("name"
+                              << "joe"
+                              << "key1" << 34 << "key2" << 45));
+        ASSERT_HAS_SPACE(st);
+    }
+}
+
+template <typename T>
+BSONObj generateSample(std::random_device& rd, T generator, size_t count) {
+    BSONObjBuilder builder;
+
+    for (size_t i = 0; i < count; ++i) {
+        builder.append("key", generator(rd));
+    }
+
+    return builder.obj();
+}
+
+// Test many metrics
+TEST(ZFTDCCompressor, TestManyMetrics) {
+    std::random_device rd;
+    std::mt19937 gen(rd());
+
+    std::uniform_int_distribution<long long> genValues(1, std::numeric_limits<long long>::max());
+    const size_t metrics = 1000;
+
+    // Test a large numbers of zeros, and incremental numbers in a full buffer
+    for (int j = 0; j < 2; j++) {
+        TestTie c;
+
+        auto st = c.addSample(generateSample(rd, genValues, metrics));
+        ASSERT_HAS_SPACE(st);
+
+        for (size_t i = 0; i <= FTDCConfig::kMaxSamplesPerArchiveMetricChunkDefault - 2; i++) {
+            st = c.addSample(generateSample(rd, genValues, metrics));
+            ASSERT_HAS_SPACE(st);
+        }
+
+        st = c.addSample(generateSample(rd, genValues, metrics));
+        ASSERT_FULL(st);
+
+        // Add Value
+        st = c.addSample(generateSample(rd, genValues, metrics));
+        ASSERT_HAS_SPACE(st);
+    }
+}
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/config.h b/src/mongo/db/ftdc/config.h
new file mode 100644
index 00000000000..ef7ee8b9143
--- /dev/null
+++ b/src/mongo/db/ftdc/config.h
@@ -0,0 +1,98 @@
+/**
+ * Copyright (C) 2015 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 <cstdint>
+
+#include "mongo/util/time_support.h"
+
+namespace mongo {
+
+/**
+ * Configuration settings for full-time diagnostic data capture (FTDC).
+ *
+ * These settings are configurable by the user at startup & runtime via setParameter.
+ */
+struct FTDCConfig {
+    FTDCConfig()
+        : enabled(kEnabledDefault),
+          maxDirectorySizeBytes(kMaxDirectorySizeBytesDefault),
+          maxFileSizeBytes(kMaxFileSizeBytesDefault),
+          period(kPeriodMillisDefault),
+          maxSamplesPerArchiveMetricChunk(kMaxSamplesPerArchiveMetricChunkDefault),
+          maxSamplesPerInterimMetricChunk(kMaxSamplesPerInterimMetricChunkDefault) {}
+
+    /**
+     * True if FTDC is collecting data. False otherwise
+     */
+    bool enabled;
+
+    /**
+     * Max Size of all FTDC files. If the total file size is > maxDirectorySizeBytes by summing up
+     * all files in the FTDC directory, the extra files are removed.
+     */
+    std::uint64_t maxDirectorySizeBytes;
+
+    /**
+     * Max size of a file in bytes.
+     */
+    std::uint64_t maxFileSizeBytes;
+
+    /**
+     * Period at which to run FTDC.
+     *
+     * FTDC is always run at the beginning at the period, and skips periods if the collector runs
+     * for more then a single period.
+     */
+    Milliseconds period;
+
+    /**
+     * Maximum number of samples to collect in an archive metric chunk for long term storage.
+     */
+    std::uint32_t maxSamplesPerArchiveMetricChunk;
+
+    /**
+     * Maximum number of samples to collect in an interim metric chunk in case the process
+     * terminates.
+     */
+    std::uint32_t maxSamplesPerInterimMetricChunk;
+
+    static const bool kEnabledDefault = true;
+
+    static const std::uint64_t kPeriodMillisDefault;
+    static const std::uint64_t kMaxDirectorySizeBytesDefault = 100 * 1024 * 1024;
+    static const std::uint64_t kMaxFileSizeBytesDefault = 10 * 1024 * 1024;
+
+    static const std::uint64_t kMaxFileUniqifier = 65000;
+
+    static const std::uint32_t kMaxSamplesPerArchiveMetricChunkDefault = 300;
+    static const std::uint32_t kMaxSamplesPerInterimMetricChunkDefault = 10;
+};
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/constants.h b/src/mongo/db/ftdc/constants.h
new file mode 100644
index 00000000000..a989fd35b64
--- /dev/null
+++ b/src/mongo/db/ftdc/constants.h
@@ -0,0 +1,45 @@
+/**
+ * Copyright (C) 2015 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.
+ */
+
+namespace mongo {
+
+extern const char kFTDCInterimFile[];
+extern const char kFTDCArchiveFile[];
+
+extern const char kFTDCIdField[];
+extern const char kFTDCTypeField[];
+
+extern const char kFTDCDataField[];
+extern const char kFTDCDocField[];
+
+extern const char kFTDCDocsField[];
+
+extern const char kFTDCCollectStartField[];
+extern const char kFTDCCollectEndField[];
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/decompressor.cpp b/src/mongo/db/ftdc/decompressor.cpp
new file mode 100644
index 00000000000..91a3d48f65c
--- /dev/null
+++ b/src/mongo/db/ftdc/decompressor.cpp
@@ -0,0 +1,169 @@
+/**
+ * Copyright (C) 2015 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/db/ftdc/decompressor.h"
+
+#include "mongo/base/data_range_cursor.h"
+#include "mongo/base/data_type_validated.h"
+#include "mongo/db/ftdc/compressor.h"
+#include "mongo/db/ftdc/util.h"
+#include "mongo/db/ftdc/varint.h"
+#include "mongo/db/jsobj.h"
+#include "mongo/util/assert_util.h"
+
+namespace mongo {
+
+StatusWith<std::vector<BSONObj>> FTDCDecompressor::uncompress(ConstDataRange buf) {
+    ConstDataRangeCursor cdc(buf);
+
+    // The document is not part of any checksum so we must validate it is correct
+    auto swRef = cdc.readAndAdvance<Validated<BSONObj>>();
+    if (!swRef.isOK()) {
+        return {swRef.getStatus()};
+    }
+
+    BSONObj ref = swRef.getValue();
+
+    // Read count of metrics
+    auto swMetricsCount = cdc.readAndAdvance<LittleEndian<std::uint32_t>>();
+    if (!swMetricsCount.isOK()) {
+        return {swMetricsCount.getStatus()};
+    }
+
+    std::uint32_t metricsCount = swMetricsCount.getValue();
+
+    // Read count of samples
+    auto swSampleCount = cdc.readAndAdvance<LittleEndian<std::uint32_t>>();
+    if (!swSampleCount.isOK()) {
+        return {swSampleCount.getStatus()};
+    }
+
+    std::uint32_t sampleCount = swSampleCount.getValue();
+
+    // Limit size of the buffer we need for metrics and samples
+    if (metricsCount * sampleCount > 1000000) {
+        return Status(ErrorCodes::InvalidLength,
+                      "Metrics Count and Sample Count have exceeded the allowable range.");
+    }
+
+    std::vector<std::uint64_t> metrics;
+
+    metrics.reserve(metricsCount);
+
+    // We pass the reference document as both the reference document and current document as we only
+    // want the array of metrics.
+    (void)FTDCBSONUtil::extractMetricsFromDocument(ref, ref, &metrics);
+
+    if (metrics.size() != metricsCount) {
+        return {ErrorCodes::BadValue,
+                "The metrics in the reference document and metrics count do not match"};
+    }
+
+    std::vector<BSONObj> docs;
+
+    // Allocate space for the reference document + samples
+    docs.reserve(1 + sampleCount);
+
+    docs.emplace_back(ref);
+
+    // We must always return the reference document
+    if (sampleCount == 0) {
+        return {docs};
+    }
+
+    // Decompress the zlib compressed buffer
+    size_t expectedDestLength = metricsCount * sampleCount * FTDCVarInt::kMaxSizeBytes64;
+
+    auto statusUncompress = _compressor.uncompress(cdc, expectedDestLength);
+
+    if (!statusUncompress.isOK()) {
+        return {statusUncompress.getStatus()};
+    }
+
+    // Read the samples
+    std::vector<std::uint64_t> deltas(metricsCount * sampleCount);
+
+    // decompress the deltas
+    std::uint64_t zeroesCount = 0;
+
+    auto cdrc = ConstDataRangeCursor(statusUncompress.getValue());
+
+    for (std::uint32_t i = 0; i < metricsCount; i++) {
+        for (std::uint32_t j = 0; j < sampleCount; j++) {
+            if (zeroesCount) {
+                deltas[FTDCCompressor::getArrayOffset(sampleCount, j, i)] = 0;
+                zeroesCount--;
+                continue;
+            }
+
+            auto swDelta = cdrc.readAndAdvance<FTDCVarInt>();
+
+            if (!swDelta.isOK()) {
+                return swDelta.getStatus();
+            }
+
+            if (swDelta.getValue() == 0) {
+                auto swZero = cdrc.readAndAdvance<FTDCVarInt>();
+
+                if (!swZero.isOK()) {
+                    return swDelta.getStatus();
+                }
+
+                zeroesCount = swZero.getValue();
+            }
+
+            deltas[FTDCCompressor::getArrayOffset(sampleCount, j, i)] = swDelta.getValue();
+        }
+    }
+
+    // Inflate the deltas
+    for (std::uint32_t i = 0; i < metricsCount; i++) {
+        deltas[FTDCCompressor::getArrayOffset(sampleCount, 0, i)] += metrics[i];
+    }
+
+    for (std::uint32_t i = 0; i < metricsCount; i++) {
+        for (std::uint32_t j = 1; j < sampleCount; j++) {
+            deltas[FTDCCompressor::getArrayOffset(sampleCount, j, i)] +=
+                deltas[FTDCCompressor::getArrayOffset(sampleCount, j - 1, i)];
+        }
+    }
+
+    for (std::uint32_t i = 0; i < sampleCount; ++i) {
+        for (std::uint32_t j = 0; j < metricsCount; ++j) {
+            metrics[j] = deltas[j * sampleCount + i];
+        }
+
+        docs.emplace_back(FTDCBSONUtil::constructDocumentFromMetrics(ref, metrics).getValue());
+    }
+
+    return {docs};
+}
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/decompressor.h b/src/mongo/db/ftdc/decompressor.h
new file mode 100644
index 00000000000..42af2985783
--- /dev/null
+++ b/src/mongo/db/ftdc/decompressor.h
@@ -0,0 +1,63 @@
+/**
+ * Copyright (C) 2015 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 <vector>
+
+#include "mongo/base/data_range.h"
+#include "mongo/base/disallow_copying.h"
+#include "mongo/base/status_with.h"
+#include "mongo/db/ftdc/block_compressor.h"
+#include "mongo/db/jsobj.h"
+
+namespace mongo {
+
+/**
+ * Inflates a compressed chunk of metrics into a list of BSON documents
+ */
+class FTDCDecompressor {
+    MONGO_DISALLOW_COPYING(FTDCDecompressor);
+
+public:
+    FTDCDecompressor() = default;
+
+    /**
+     * Inflates a compressed chunk of metrics into a vector of owned BSON documents.
+     *
+     * Will fail if the chunk is corrupt or too short.
+     *
+     * Returns N samples where N = sample count + 1. The 1 is the reference document.
+     */
+    StatusWith<std::vector<BSONObj>> uncompress(ConstDataRange buf);
+
+private:
+    BlockCompressor _compressor;
+};
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/util.cpp b/src/mongo/db/ftdc/util.cpp
new file mode 100644
index 00000000000..a154fb92db5
--- /dev/null
+++ b/src/mongo/db/ftdc/util.cpp
@@ -0,0 +1,399 @@
+/**
+ * Copyright (C) 2015 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::kFTDC
+
+#include "mongo/platform/basic.h"
+
+#include "mongo/db/ftdc/util.h"
+
+#include <boost/filesystem.hpp>
+
+#include "mongo/bson/bsonobjbuilder.h"
+#include "mongo/bson/util/bson_extract.h"
+#include "mongo/config.h"
+#include "mongo/db/ftdc/config.h"
+#include "mongo/db/ftdc/constants.h"
+#include "mongo/db/jsobj.h"
+#include "mongo/db/service_context.h"
+#include "mongo/util/assert_util.h"
+#include "mongo/util/log.h"
+#include "mongo/util/mongoutils/str.h"
+#include "mongo/util/time_support.h"
+
+namespace mongo {
+
+const char kFTDCInterimFile[] = "metrics.interim";
+const char kFTDCArchiveFile[] = "metrics";
+
+const char kFTDCIdField[] = "_id";
+const char kFTDCTypeField[] = "type";
+
+const char kFTDCDataField[] = "data";
+const char kFTDCDocField[] = "doc";
+
+const char kFTDCDocsField[] = "docs";
+
+const char kFTDCCollectStartField[] = "start";
+const char kFTDCCollectEndField[] = "end";
+
+const std::uint64_t FTDCConfig::kPeriodMillisDefault = 1000;
+
+const std::size_t kMaxRecursion = 10;
+
+namespace FTDCUtil {
+
+boost::filesystem::path getInterimFile(const boost::filesystem::path& file) {
+    if (boost::filesystem::is_directory(file)) {
+        return file / kFTDCInterimFile;
+    }
+
+    auto p = file.parent_path();
+    p /= kFTDCInterimFile;
+
+    return p;
+}
+
+Date_t roundTime(Date_t now, Milliseconds period) {
+    // Note: auto type deduction is explicitly avoided here to ensure rigid type correctness
+    long long clock_duration = now.toMillisSinceEpoch();
+
+    long long now_next_period = clock_duration + period.count();
+
+    long long excess_time(now_next_period % period.count());
+
+    long long next_time = now_next_period - excess_time;
+
+    return Date_t::fromMillisSinceEpoch(next_time);
+}
+
+}  // namespace FTDCUtil
+
+
+namespace FTDCBSONUtil {
+
+namespace {
+
+StatusWith<bool> extractMetricsFromDocument(const BSONObj& referenceDoc,
+                                            const BSONObj& currentDoc,
+                                            std::vector<std::uint64_t>* metrics,
+                                            size_t recursion) {
+    if (recursion > kMaxRecursion) {
+        return {ErrorCodes::BadValue, "Recursion limit reached."};
+    }
+
+    bool matches = true;
+    BSONObjIterator itCurrent(currentDoc);
+    BSONObjIterator itReference(referenceDoc);
+
+    while (itCurrent.more()) {
+        // Schema mismatch if current document is longer than reference document
+        if (matches && !itReference.more()) {
+            LOG(4) << "full-time diagnostic data capture schema change: currrent document is "
+                      "longer than "
+                      "reference document";
+            matches = false;
+        }
+
+        BSONElement currentElement = itCurrent.next();
+        BSONElement referenceElement = matches ? itReference.next() : BSONElement();
+
+        if (matches) {
+            // Check for matching field names
+            if (referenceElement.fieldNameStringData() != currentElement.fieldNameStringData()) {
+                LOG(4)
+                    << "full-time diagnostic data capture schema change: field name change - from '"
+                    << referenceElement.fieldNameStringData() << "' to '"
+                    << currentElement.fieldNameStringData() << "'";
+                matches = false;
+            }
+
+            // Check that types match, allowing any numeric type to match any other numeric type.
+            // This looseness is necessary because some metrics use varying numeric types,
+            // and if that was considered a schema mismatch, it would increase the number of
+            // reference samples required.
+            if ((currentElement.type() != referenceElement.type()) &&
+                !(referenceElement.isNumber() == true &&
+                  currentElement.isNumber() == referenceElement.isNumber())) {
+                LOG(4) << "full-time diagnostic data capture  schema change: field type change for "
+                          "field '" << referenceElement.fieldNameStringData() << "' from '"
+                       << static_cast<int>(referenceElement.type()) << "' to '"
+                       << static_cast<int>(currentElement.type()) << "'";
+                matches = false;
+            }
+        }
+
+        switch (currentElement.type()) {
+            // all numeric types are extracted as long (int64)
+            // this supports the loose schema matching mentioned above,
+            // but does create a range issue for doubles, and requires doubles to be integer
+            case NumberDouble:
+            case NumberInt:
+            case NumberLong:
+            case NumberDecimal:
+                metrics->emplace_back(currentElement.numberLong());
+                break;
+
+            case Bool:
+                metrics->emplace_back(currentElement.Bool());
+                break;
+
+            case Date:
+                metrics->emplace_back(currentElement.Date().toMillisSinceEpoch());
+                break;
+
+            case bsonTimestamp:
+                // very slightly more space efficient to treat these as two separate metrics
+                metrics->emplace_back(currentElement.timestampValue());
+                metrics->emplace_back(currentElement.timestampInc());
+                break;
+
+            case Object:
+            case Array: {
+                // Maximum recursion is controlled by the documents we collect. Maximum is 5 in the
+                // current implementation.
+                auto sw = extractMetricsFromDocument(matches ? referenceElement.Obj() : BSONObj(),
+                                                     currentElement.Obj(),
+                                                     metrics,
+                                                     recursion + 1);
+                if (!sw.isOK()) {
+                    return sw;
+                }
+                matches = matches && sw.getValue();
+            } break;
+
+            default:
+                break;
+        }
+    }
+
+    // schema mismatch if ref is longer than curr
+    if (matches && itReference.more()) {
+        LOG(4) << "full-time diagnostic data capture schema change: reference document is longer "
+                  "then current";
+        matches = false;
+    }
+
+    return {matches};
+}
+
+}  // namespace
+
+StatusWith<bool> extractMetricsFromDocument(const BSONObj& referenceDoc,
+                                            const BSONObj& currentDoc,
+                                            std::vector<std::uint64_t>* metrics) {
+    return extractMetricsFromDocument(referenceDoc, currentDoc, metrics, 0);
+}
+
+namespace {
+Status constructDocumentFromMetrics(const BSONObj& referenceDocument,
+                                    BSONObjBuilder& builder,
+                                    const std::vector<std::uint64_t>& metrics,
+                                    size_t* pos,
+                                    size_t recursion) {
+    if (recursion > kMaxRecursion) {
+        return {ErrorCodes::BadValue, "Recursion limit reached."};
+    }
+
+    BSONObjIterator iterator(referenceDocument);
+    while (iterator.more()) {
+        BSONElement currentElement = iterator.next();
+
+        switch (currentElement.type()) {
+            case NumberDouble:
+            case NumberInt:
+            case NumberLong:
+            case NumberDecimal:
+                if (*pos >= metrics.size()) {
+                    return Status(
+                        ErrorCodes::BadValue,
+                        "There are more metrics in the reference document then expected.");
+                }
+
+                builder.append(currentElement.fieldName(),
+                               static_cast<long long int>(metrics[(*pos)++]));
+                break;
+
+            case Bool:
+                if (*pos >= metrics.size()) {
+                    return Status(
+                        ErrorCodes::BadValue,
+                        "There are more metrics in the reference document then expected.");
+                }
+
+                builder.append(currentElement.fieldName(), static_cast<bool>(metrics[(*pos)++]));
+                break;
+
+            case Date:
+                if (*pos >= metrics.size()) {
+                    return Status(
+                        ErrorCodes::BadValue,
+                        "There are more metrics in the reference document then expected.");
+                }
+
+                builder.append(
+                    currentElement.fieldName(),
+                    Date_t::fromMillisSinceEpoch(static_cast<std::uint64_t>(metrics[(*pos)++])));
+                break;
+
+            case bsonTimestamp: {
+                if (*pos + 1 >= metrics.size()) {
+                    return Status(
+                        ErrorCodes::BadValue,
+                        "There are more metrics in the reference document then expected.");
+                }
+
+                std::uint64_t seconds = metrics[(*pos)++];
+                std::uint64_t increment = metrics[(*pos)++];
+                builder.append(currentElement.fieldName(), Timestamp(seconds, increment));
+                break;
+            }
+
+            case Object: {
+                BSONObjBuilder sub(builder.subobjStart(currentElement.fieldName()));
+                auto s = constructDocumentFromMetrics(
+                    currentElement.Obj(), sub, metrics, pos, recursion + 1);
+                if (!s.isOK()) {
+                    return s;
+                }
+                break;
+            }
+
+            case Array: {
+                BSONObjBuilder sub(builder.subarrayStart(currentElement.fieldName()));
+                auto s = constructDocumentFromMetrics(
+                    currentElement.Obj(), sub, metrics, pos, recursion + 1);
+                if (!s.isOK()) {
+                    return s;
+                }
+                break;
+            }
+
+            default:
+                builder.append(currentElement);
+                break;
+        }
+    }
+
+    return Status::OK();
+}
+
+}  // namespace
+
+StatusWith<BSONObj> constructDocumentFromMetrics(const BSONObj& ref,
+                                                 const std::vector<std::uint64_t>& metrics) {
+    size_t at = 0;
+    BSONObjBuilder b;
+    Status s = constructDocumentFromMetrics(ref, b, metrics, &at, 0);
+    if (!s.isOK()) {
+        return StatusWith<BSONObj>(s);
+    }
+
+    return b.obj();
+}
+
+BSONObj createBSONMetadataDocument(const BSONObj& metadata) {
+    BSONObjBuilder builder;
+    builder.appendDate(kFTDCIdField, getGlobalServiceContext()->getClockSource()->now());
+    builder.appendNumber(kFTDCTypeField, static_cast<int>(FTDCType::kMetadata));
+    builder.appendObject(kFTDCDocField, metadata.objdata(), metadata.objsize());
+
+    return builder.obj();
+}
+
+BSONObj createBSONMetricChunkDocument(ConstDataRange buf) {
+    BSONObjBuilder builder;
+
+    builder.appendDate(kFTDCIdField, getGlobalServiceContext()->getClockSource()->now());
+    builder.appendNumber(kFTDCTypeField, static_cast<int>(FTDCType::kMetricChunk));
+    builder.appendBinData(kFTDCDataField, buf.length(), BinDataType::BinDataGeneral, buf.data());
+
+    return builder.obj();
+}
+
+StatusWith<FTDCType> getBSONDocumentType(const BSONObj& obj) {
+    long long value;
+
+    Status status = bsonExtractIntegerField(obj, kFTDCTypeField, &value);
+    if (!status.isOK()) {
+        return {status};
+    }
+
+    if (static_cast<FTDCType>(value) != FTDCType::kMetricChunk &&
+        static_cast<FTDCType>(value) != FTDCType::kMetadata) {
+        return {ErrorCodes::BadValue,
+                str::stream() << "Field '" << std::string(kFTDCTypeField)
+                              << "' is not an expected value, found '" << value << "'"};
+    }
+
+    return {static_cast<FTDCType>(value)};
+}
+
+StatusWith<BSONObj> getBSONDocumentFromMetadataDoc(const BSONObj& obj) {
+    if (kDebugBuild) {
+        auto swType = getBSONDocumentType(obj);
+        dassert(swType.isOK() && swType.getValue() == FTDCType::kMetadata);
+    }
+
+    BSONElement element;
+
+    Status status = bsonExtractTypedField(obj, kFTDCDocField, BSONType::Object, &element);
+    if (!status.isOK()) {
+        return {status};
+    }
+
+    return {element.Obj()};
+}
+
+StatusWith<std::vector<BSONObj>> getMetricsFromMetricDoc(const BSONObj& obj,
+                                                         FTDCDecompressor* decompressor) {
+    if (kDebugBuild) {
+        auto swType = getBSONDocumentType(obj);
+        dassert(swType.isOK() && swType.getValue() == FTDCType::kMetricChunk);
+    }
+
+    BSONElement element;
+
+    Status status = bsonExtractTypedField(obj, kFTDCDataField, BSONType::BinData, &element);
+    if (!status.isOK()) {
+        return {status};
+    }
+
+    int length;
+    const char* buffer = element.binData(length);
+    if (length < 0) {
+        return {ErrorCodes::BadValue,
+                str::stream() << "Field " << std::string(kFTDCTypeField) << " is not a BinData."};
+    }
+
+    return decompressor->uncompress({buffer, static_cast<std::size_t>(length)});
+}
+
+}  // namespace FTDCBSONUtil
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/util.h b/src/mongo/db/ftdc/util.h
new file mode 100644
index 00000000000..f2b57b77a81
--- /dev/null
+++ b/src/mongo/db/ftdc/util.h
@@ -0,0 +1,184 @@
+/**
+ * Copyright (C) 2015 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/filesystem/path.hpp>
+#include <vector>
+
+#include "mongo/base/status.h"
+#include "mongo/base/status_with.h"
+#include "mongo/db/ftdc/decompressor.h"
+#include "mongo/db/jsobj.h"
+
+namespace mongo {
+
+/**
+ * Utilities for inflating and deflating BSON documents and metric arrays
+ */
+namespace FTDCBSONUtil {
+
+/**
+* Type of FTDC document.
+*
+* NOTE: Persisted to disk via BSON Objects.
+*/
+enum class FTDCType : std::int32_t {
+    /**
+    * A metadata document is composed of a header + an array of bson documents
+    *
+    * See createBSONMetadataChunkDocument
+    */
+    kMetadata = 0,
+
+    /**
+    * A metrics chunk is composed of a header + a compressed metric chunk.
+    *
+    * See createBSONMetricChunkDocument
+    */
+    kMetricChunk = 1,
+};
+
+
+/**
+ * Extract an array of numbers from a pair of documents. Verifies the pair of documents have same
+ * structure.
+ *
+ * Types considered numbers for the purpose of metrics:
+ *  - double - encoded as an integer, loses fractional components via truncation
+ *  - 32-bit integer
+ *  - 64-integer
+ *  - bool
+ *  - date
+ *  - timestamp
+ *    Note: Timestamp is encoded as two integers, the timestamp value followed by the increment.
+ *
+ * Two documents are considered the same if satisfy the following criteria:
+ *
+ * Criteria: During a Depth First traversal of the document:
+ *  1. Each element has the same name regardless of its type.
+ *  2. The same number of elements exist in each document.
+ *  3. The types of each element are the same.
+ *     Note: Double, Int, and Long are treated as equivalent for this purpose.
+ *
+ * @param referenceDoc A reference document to use the as the definition of the correct schema.
+ * @param doc A second document to compare against the reference document and extract metrics
+ * from
+ * @param metrics A vector of metrics that were extracted from the doc
+ *
+ * \return false if the documents differ in terms of metrics
+ */
+StatusWith<bool> extractMetricsFromDocument(const BSONObj& referenceDoc,
+                                            const BSONObj& doc,
+                                            std::vector<std::uint64_t>* metrics);
+
+/**
+ * Construct a document from a reference document and array of metrics.
+ *
+ * @param referenceDoc A reference document to use the as the definition of the correct schema.
+ * @param builder A BSON builder to construct a single document into. Each document will be a
+ *copy
+ * of the reference document with the numerical fields replaced with values from metrics array.
+ * @param metrics A vector of metrics for all documents
+ * @param pos A position into the array of metrics to start iterating from.
+ *
+ * \return Status if the decompression of the buffer was successful or failed. Decompression may
+ * fail if the buffer is not valid.
+ */
+Status constructDocumentFromMetrics(const BSONObj& referenceDoc,
+                                    BSONObjBuilder& builder,
+                                    const std::vector<std::uint64_t>& metrics,
+                                    size_t* pos);
+
+/**
+ * Construct a document from a reference document and array of metrics. See documentation above.
+ */
+StatusWith<BSONObj> constructDocumentFromMetrics(const BSONObj& ref,
+                                                 const std::vector<std::uint64_t>& metrics);
+
+/**
+ * Create BSON metadata document for storage. The passed in document is embedded as the doc
+ * field in the example above.
+ *
+ * Example:
+ * {
+ *  "_id" : Date_t
+ *  "type" : 0
+ *  "doc" : { ... }
+ * }
+ */
+BSONObj createBSONMetadataDocument(const BSONObj& metadata);
+
+/**
+ * Create a BSON metric chunk document for storage. The passed in document is embedded as the
+ * data field in the example above.
+ *
+ * Example:
+ * {
+ *  "_id" : Date_t
+ *  "type" : 1
+ *  "data" : BinData(...)
+ * }
+ */
+BSONObj createBSONMetricChunkDocument(ConstDataRange buf);
+
+/**
+ * Get the type of a BSON document
+ */
+StatusWith<FTDCType> getBSONDocumentType(const BSONObj& obj);
+
+/**
+ * Extract the metadata field from a BSON document
+ */
+StatusWith<BSONObj> getBSONDocumentFromMetadataDoc(const BSONObj& obj);
+
+/**
+ * Get the set of metric documents from the compressed chunk of a metric document
+ */
+StatusWith<std::vector<BSONObj>> getMetricsFromMetricDoc(const BSONObj& obj,
+                                                         FTDCDecompressor* decompressor);
+}  // namespace FTDCBSONUtil
+
+
+/**
+ * Miscellaneous utilties for FTDC.
+ */
+namespace FTDCUtil {
+/**
+ * Construct the full path to the interim file
+ */
+boost::filesystem::path getInterimFile(const boost::filesystem::path& file);
+
+/**
+ * Round the specified time_point to the next multiple of period after the specified time_point
+ */
+Date_t roundTime(Date_t now, Milliseconds period);
+
+}  // namespace FTDCUtil
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/util_test.cpp b/src/mongo/db/ftdc/util_test.cpp
new file mode 100644
index 00000000000..aa99006c73a
--- /dev/null
+++ b/src/mongo/db/ftdc/util_test.cpp
@@ -0,0 +1,49 @@
+/**
+ * Copyright (C) 2015 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/db/ftdc/util.h"
+#include "mongo/unittest/unittest.h"
+
+namespace mongo {
+
+void checkTime(int expected, int now_time, int period) {
+    ASSERT_TRUE(Date_t::fromMillisSinceEpoch(expected) ==
+                FTDCUtil::roundTime(Date_t::fromMillisSinceEpoch(now_time), Milliseconds(period)));
+}
+
+// Validate time rounding
+TEST(FTDCUtilTest, TestRoundTime) {
+    checkTime(4, 3, 1);
+    checkTime(7, 3, 7);
+    checkTime(14, 8, 7);
+    checkTime(14, 13, 7);
+}
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/varint.cpp b/src/mongo/db/ftdc/varint.cpp
new file mode 100644
index 00000000000..d4b0a8819c4
--- /dev/null
+++ b/src/mongo/db/ftdc/varint.cpp
@@ -0,0 +1,84 @@
+/**
+ * Copyright (C) 2015 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/db/ftdc/varint.h"
+
+#include <third_party/s2/util/coding/varint.h>
+
+#include "mongo/util/assert_util.h"
+
+namespace mongo {
+
+Status DataType::Handler<FTDCVarInt>::load(
+    FTDCVarInt* t, const char* ptr, size_t length, size_t* advanced, std::ptrdiff_t debug_offset) {
+    std::uint64_t value;
+
+    const char* newptr =
+        Varint::Parse64WithLimit(ptr, ptr + length, reinterpret_cast<uint64*>(&value));
+
+    if (!newptr) {
+        return DataType::makeTrivialLoadStatus(FTDCVarInt::kMaxSizeBytes64, length, debug_offset);
+    }
+
+    if (t) {
+        *t = value;
+    }
+
+    if (advanced) {
+        *advanced = newptr - ptr;
+    }
+
+    return Status::OK();
+}
+
+Status DataType::Handler<FTDCVarInt>::store(
+    const FTDCVarInt& t, char* ptr, size_t length, size_t* advanced, std::ptrdiff_t debug_offset) {
+    // nullptr means it wants to know how much space we want
+    if (!ptr) {
+        *advanced = FTDCVarInt::kMaxSizeBytes64;
+        return Status::OK();
+    }
+
+    if (FTDCVarInt::kMaxSizeBytes64 > length) {
+        return DataType::makeTrivialStoreStatus(FTDCVarInt::kMaxSizeBytes64, length, debug_offset);
+    }
+
+    // Use a dassert since static_assert does not work because the expression does not have a
+    // constant value
+    dassert(Varint::kMax64 == FTDCVarInt::kMaxSizeBytes64);
+
+    const char* newptr = Varint::Encode64(ptr, t);
+
+    if (advanced) {
+        *advanced = newptr - ptr;
+    }
+
+    return Status::OK();
+}
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/varint.h b/src/mongo/db/ftdc/varint.h
new file mode 100644
index 00000000000..0dd4c73fb1b
--- /dev/null
+++ b/src/mongo/db/ftdc/varint.h
@@ -0,0 +1,92 @@
+/**
+ * Copyright (C) 2015 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 <cstddef>
+#include <cstdint>
+
+#include "mongo/base/status.h"
+#include "mongo/base/data_type.h"
+
+namespace mongo {
+/**
+ * Methods to compress and decompress 64-bit integers into variable integers
+ *
+ * Uses a technique described here:
+ * S. Buttcher, C. L. A. Clarke, and G. V. Cormack.
+ *  Information Retrieval: Implementing and Evaluating Search Engines. MIT Press, Cambridge, MA,
+ *  2010
+ */
+struct FTDCVarInt {
+    /**
+    * Maximum number of bytes an integer can compress to
+    */
+    static const std::size_t kMaxSizeBytes64 = 10;
+
+    FTDCVarInt() = default;
+    FTDCVarInt(std::uint64_t t) : _value(t) {}
+
+    operator std::uint64_t() const {
+        return _value;
+    }
+
+private:
+    std::uint64_t _value{0};
+};
+
+template <>
+struct DataType::Handler<FTDCVarInt> {
+    /**
+     * Compress a 64-bit integer and return the new buffer position.
+     *
+     * end should be the byte after the end of the buffer.
+     *
+     * Return nullptr for bad encoded data.
+     */
+    static Status load(FTDCVarInt* t,
+                       const char* ptr,
+                       size_t length,
+                       size_t* advanced,
+                       std::ptrdiff_t debug_offset);
+
+    /**
+     * Compress a 64-bit integer and return the new buffer position
+     */
+    static Status store(const FTDCVarInt& t,
+                        char* ptr,
+                        size_t length,
+                        size_t* advanced,
+                        std::ptrdiff_t debug_offset);
+
+    static FTDCVarInt defaultConstruct() {
+        return 0;
+    }
+};
+
+}  // namespace mongo
diff --git a/src/mongo/db/ftdc/varint_test.cpp b/src/mongo/db/ftdc/varint_test.cpp
new file mode 100644
index 00000000000..cb5f303981c
--- /dev/null
+++ b/src/mongo/db/ftdc/varint_test.cpp
@@ -0,0 +1,88 @@
+/**
+ * Copyright (C) 2015 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/base/data_builder.h"
+#include "mongo/base/data_type_validated.h"
+#include "mongo/base/init.h"
+#include "mongo/db/client.h"
+#include "mongo/db/ftdc/collector.h"
+#include "mongo/db/ftdc/config.h"
+#include "mongo/db/ftdc/controller.h"
+#include "mongo/db/ftdc/varint.h"
+#include "mongo/unittest/unittest.h"
+
+namespace mongo {
+
+// Test integer packing and unpacking
+void TestInt(std::uint64_t i) {
+    char buf[11];
+
+    DataView dvWrite(&buf[0]);
+
+    dvWrite.write(i);
+
+    ConstDataView cdvRead(&buf[0]);
+
+    std::uint64_t d = cdvRead.read<std::uint64_t>();
+
+    ASSERT_EQUALS(i, d);
+}
+
+// Test various integer combinations compress and uncompress correctly
+TEST(FTDCVarIntTest, TestIntCompression) {
+    // Check numbers with leading 1
+    for (int i = 0; i < 63; i++) {
+        TestInt(i);
+        TestInt(i - 1);
+    }
+
+    // Check numbers composed of repeating hex numbers
+    for (int i = 0; i < 15; i++) {
+        int v = 0;
+        for (int j = 0; j < 15; j++) {
+            v = v << 4 | i;
+            TestInt(v);
+        }
+    }
+}
+
+// Test data builder can write a lot of zeros
+TEST(FTDCVarIntTest, TestDataBuilder) {
+    DataBuilder db(1);
+
+    // DataBuilder grows by 2x, and we reserve 10 bytes
+    // lcm(2**x, 10) == 16
+    for (int i = 0; i < 16; i++) {
+        auto s1 = db.writeAndAdvance(FTDCVarInt(0));
+        ASSERT_OK(s1);
+    };
+}
+
+}  // namespace mongo