rocksdb/table/merger.cc

361 lines
10 KiB
C++
Raw Normal View History

// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
//
// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
#include "table/merger.h"
#include <vector>
#include "db/pinned_iterators_manager.h"
#include "rocksdb/comparator.h"
#include "rocksdb/iterator.h"
#include "rocksdb/options.h"
#include "table/internal_iterator.h"
#include "table/iter_heap.h"
#include "table/iterator_wrapper.h"
#include "util/arena.h"
#include "util/autovector.h"
#include "util/heap.h"
#include "util/perf_context_imp.h"
#include "util/stop_watch.h"
#include "util/sync_point.h"
namespace rocksdb {
// Without anonymous namespace here, we fail the warning -Wmissing-prototypes
namespace {
typedef BinaryHeap<IteratorWrapper*, MaxIteratorComparator> MergerMaxIterHeap;
typedef BinaryHeap<IteratorWrapper*, MinIteratorComparator> MergerMinIterHeap;
} // namespace
const size_t kNumIterReserve = 4;
class MergingIterator : public InternalIterator {
public:
MergingIterator(const Comparator* comparator, InternalIterator** children,
int n, bool is_arena_mode)
: is_arena_mode_(is_arena_mode),
comparator_(comparator),
current_(nullptr),
direction_(kForward),
minHeap_(comparator_),
pinned_iters_mgr_(nullptr) {
children_.resize(n);
for (int i = 0; i < n; i++) {
children_[i].Set(children[i]);
}
for (auto& child : children_) {
if (child.Valid()) {
minHeap_.push(&child);
}
}
current_ = CurrentForward();
}
virtual void AddIterator(InternalIterator* iter) {
assert(direction_ == kForward);
children_.emplace_back(iter);
if (pinned_iters_mgr_) {
iter->SetPinnedItersMgr(pinned_iters_mgr_);
Introduce ReadOptions::pin_data (support zero copy for keys) Summary: This patch update the Iterator API to introduce new functions that allow users to keep the Slices returned by key() valid as long as the Iterator is not deleted ReadOptions::pin_data : If true keep loaded blocks in memory as long as the iterator is not deleted Iterator::IsKeyPinned() : If true, this mean that the Slice returned by key() is valid as long as the iterator is not deleted Also add a new option BlockBasedTableOptions::use_delta_encoding to allow users to disable delta_encoding if needed. Benchmark results (using https://phabricator.fb.com/P20083553) ``` // $ du -h /home/tec/local/normal.4K.Snappy/db10077 // 6.1G /home/tec/local/normal.4K.Snappy/db10077 // $ du -h /home/tec/local/zero.8K.LZ4/db10077 // 6.4G /home/tec/local/zero.8K.LZ4/db10077 // Benchmarks for shard db10077 // _build/opt/rocks/benchmark/rocks_copy_benchmark \ // --normal_db_path="/home/tec/local/normal.4K.Snappy/db10077" \ // --zero_db_path="/home/tec/local/zero.8K.LZ4/db10077" // First run // ============================================================================ // rocks/benchmark/RocksCopyBenchmark.cpp relative time/iter iters/s // ============================================================================ // BM_StringCopy 1.73s 576.97m // BM_StringPiece 103.74% 1.67s 598.55m // ============================================================================ // Match rate : 1000000 / 1000000 // Second run // ============================================================================ // rocks/benchmark/RocksCopyBenchmark.cpp relative time/iter iters/s // ============================================================================ // BM_StringCopy 611.99ms 1.63 // BM_StringPiece 203.76% 300.35ms 3.33 // ============================================================================ // Match rate : 1000000 / 1000000 ``` Test Plan: Unit tests Reviewers: sdong, igor, anthony, yhchiang, rven Reviewed By: rven Subscribers: dhruba, lovro, adsharma Differential Revision: https://reviews.facebook.net/D48999
2015-12-16 20:08:30 +00:00
}
auto new_wrapper = children_.back();
if (new_wrapper.Valid()) {
minHeap_.push(&new_wrapper);
current_ = CurrentForward();
}
}
virtual ~MergingIterator() {
for (auto& child : children_) {
child.DeleteIter(is_arena_mode_);
}
}
virtual bool Valid() const override { return (current_ != nullptr); }
virtual void SeekToFirst() override {
ClearHeaps();
for (auto& child : children_) {
child.SeekToFirst();
if (child.Valid()) {
minHeap_.push(&child);
}
}
direction_ = kForward;
current_ = CurrentForward();
}
virtual void SeekToLast() override {
ClearHeaps();
InitMaxHeap();
for (auto& child : children_) {
child.SeekToLast();
if (child.Valid()) {
maxHeap_->push(&child);
}
}
direction_ = kReverse;
current_ = CurrentReverse();
}
virtual void Seek(const Slice& target) override {
ClearHeaps();
for (auto& child : children_) {
{
PERF_TIMER_GUARD(seek_child_seek_time);
child.Seek(target);
}
PERF_COUNTER_ADD(seek_child_seek_count, 1);
if (child.Valid()) {
PERF_TIMER_GUARD(seek_min_heap_time);
minHeap_.push(&child);
}
}
direction_ = kForward;
{
PERF_TIMER_GUARD(seek_min_heap_time);
current_ = CurrentForward();
}
}
virtual void Next() override {
assert(Valid());
// Ensure that all children are positioned after key().
// If we are moving in the forward direction, it is already
// true for all of the non-current children since current_ is
// the smallest child and key() == current_->key().
if (direction_ != kForward) {
// Otherwise, advance the non-current children. We advance current_
// just after the if-block.
ClearHeaps();
for (auto& child : children_) {
if (&child != current_) {
child.Seek(key());
if (child.Valid() && comparator_->Equal(key(), child.key())) {
child.Next();
}
}
if (child.Valid()) {
minHeap_.push(&child);
}
}
direction_ = kForward;
// The loop advanced all non-current children to be > key() so current_
// should still be strictly the smallest key.
assert(current_ == CurrentForward());
}
// For the heap modifications below to be correct, current_ must be the
// current top of the heap.
assert(current_ == CurrentForward());
// as the current points to the current record. move the iterator forward.
current_->Next();
if (current_->Valid()) {
// current is still valid after the Next() call above. Call
// replace_top() to restore the heap property. When the same child
// iterator yields a sequence of keys, this is cheap.
minHeap_.replace_top(current_);
} else {
// current stopped being valid, remove it from the heap.
minHeap_.pop();
}
current_ = CurrentForward();
}
virtual void Prev() override {
assert(Valid());
// Ensure that all children are positioned before key().
// If we are moving in the reverse direction, it is already
// true for all of the non-current children since current_ is
// the largest child and key() == current_->key().
if (direction_ != kReverse) {
// Otherwise, retreat the non-current children. We retreat current_
// just after the if-block.
ClearHeaps();
InitMaxHeap();
for (auto& child : children_) {
if (&child != current_) {
child.Seek(key());
if (child.Valid()) {
// Child is at first entry >= key(). Step back one to be < key()
TEST_SYNC_POINT_CALLBACK("MergeIterator::Prev:BeforePrev", &child);
child.Prev();
} else {
// Child has no entries >= key(). Position at last entry.
TEST_SYNC_POINT("MergeIterator::Prev:BeforeSeekToLast");
child.SeekToLast();
}
}
if (child.Valid()) {
maxHeap_->push(&child);
}
}
direction_ = kReverse;
// Note that we don't do assert(current_ == CurrentReverse()) here
// because it is possible to have some keys larger than the seek-key
// inserted between Seek() and SeekToLast(), which makes current_ not
// equal to CurrentReverse().
current_ = CurrentReverse();
}
// For the heap modifications below to be correct, current_ must be the
// current top of the heap.
assert(current_ == CurrentReverse());
current_->Prev();
if (current_->Valid()) {
// current is still valid after the Prev() call above. Call
// replace_top() to restore the heap property. When the same child
// iterator yields a sequence of keys, this is cheap.
maxHeap_->replace_top(current_);
} else {
// current stopped being valid, remove it from the heap.
maxHeap_->pop();
}
current_ = CurrentReverse();
}
virtual Slice key() const override {
assert(Valid());
return current_->key();
}
virtual Slice value() const override {
assert(Valid());
return current_->value();
}
virtual Status status() const override {
Status s;
for (auto& child : children_) {
s = child.status();
if (!s.ok()) {
break;
}
}
return s;
}
virtual void SetPinnedItersMgr(
PinnedIteratorsManager* pinned_iters_mgr) override {
pinned_iters_mgr_ = pinned_iters_mgr;
Introduce ReadOptions::pin_data (support zero copy for keys) Summary: This patch update the Iterator API to introduce new functions that allow users to keep the Slices returned by key() valid as long as the Iterator is not deleted ReadOptions::pin_data : If true keep loaded blocks in memory as long as the iterator is not deleted Iterator::IsKeyPinned() : If true, this mean that the Slice returned by key() is valid as long as the iterator is not deleted Also add a new option BlockBasedTableOptions::use_delta_encoding to allow users to disable delta_encoding if needed. Benchmark results (using https://phabricator.fb.com/P20083553) ``` // $ du -h /home/tec/local/normal.4K.Snappy/db10077 // 6.1G /home/tec/local/normal.4K.Snappy/db10077 // $ du -h /home/tec/local/zero.8K.LZ4/db10077 // 6.4G /home/tec/local/zero.8K.LZ4/db10077 // Benchmarks for shard db10077 // _build/opt/rocks/benchmark/rocks_copy_benchmark \ // --normal_db_path="/home/tec/local/normal.4K.Snappy/db10077" \ // --zero_db_path="/home/tec/local/zero.8K.LZ4/db10077" // First run // ============================================================================ // rocks/benchmark/RocksCopyBenchmark.cpp relative time/iter iters/s // ============================================================================ // BM_StringCopy 1.73s 576.97m // BM_StringPiece 103.74% 1.67s 598.55m // ============================================================================ // Match rate : 1000000 / 1000000 // Second run // ============================================================================ // rocks/benchmark/RocksCopyBenchmark.cpp relative time/iter iters/s // ============================================================================ // BM_StringCopy 611.99ms 1.63 // BM_StringPiece 203.76% 300.35ms 3.33 // ============================================================================ // Match rate : 1000000 / 1000000 ``` Test Plan: Unit tests Reviewers: sdong, igor, anthony, yhchiang, rven Reviewed By: rven Subscribers: dhruba, lovro, adsharma Differential Revision: https://reviews.facebook.net/D48999
2015-12-16 20:08:30 +00:00
for (auto& child : children_) {
child.SetPinnedItersMgr(pinned_iters_mgr);
Introduce ReadOptions::pin_data (support zero copy for keys) Summary: This patch update the Iterator API to introduce new functions that allow users to keep the Slices returned by key() valid as long as the Iterator is not deleted ReadOptions::pin_data : If true keep loaded blocks in memory as long as the iterator is not deleted Iterator::IsKeyPinned() : If true, this mean that the Slice returned by key() is valid as long as the iterator is not deleted Also add a new option BlockBasedTableOptions::use_delta_encoding to allow users to disable delta_encoding if needed. Benchmark results (using https://phabricator.fb.com/P20083553) ``` // $ du -h /home/tec/local/normal.4K.Snappy/db10077 // 6.1G /home/tec/local/normal.4K.Snappy/db10077 // $ du -h /home/tec/local/zero.8K.LZ4/db10077 // 6.4G /home/tec/local/zero.8K.LZ4/db10077 // Benchmarks for shard db10077 // _build/opt/rocks/benchmark/rocks_copy_benchmark \ // --normal_db_path="/home/tec/local/normal.4K.Snappy/db10077" \ // --zero_db_path="/home/tec/local/zero.8K.LZ4/db10077" // First run // ============================================================================ // rocks/benchmark/RocksCopyBenchmark.cpp relative time/iter iters/s // ============================================================================ // BM_StringCopy 1.73s 576.97m // BM_StringPiece 103.74% 1.67s 598.55m // ============================================================================ // Match rate : 1000000 / 1000000 // Second run // ============================================================================ // rocks/benchmark/RocksCopyBenchmark.cpp relative time/iter iters/s // ============================================================================ // BM_StringCopy 611.99ms 1.63 // BM_StringPiece 203.76% 300.35ms 3.33 // ============================================================================ // Match rate : 1000000 / 1000000 ``` Test Plan: Unit tests Reviewers: sdong, igor, anthony, yhchiang, rven Reviewed By: rven Subscribers: dhruba, lovro, adsharma Differential Revision: https://reviews.facebook.net/D48999
2015-12-16 20:08:30 +00:00
}
}
virtual bool IsKeyPinned() const override {
assert(Valid());
return pinned_iters_mgr_ && pinned_iters_mgr_->PinningEnabled() &&
current_->IsKeyPinned();
Introduce ReadOptions::pin_data (support zero copy for keys) Summary: This patch update the Iterator API to introduce new functions that allow users to keep the Slices returned by key() valid as long as the Iterator is not deleted ReadOptions::pin_data : If true keep loaded blocks in memory as long as the iterator is not deleted Iterator::IsKeyPinned() : If true, this mean that the Slice returned by key() is valid as long as the iterator is not deleted Also add a new option BlockBasedTableOptions::use_delta_encoding to allow users to disable delta_encoding if needed. Benchmark results (using https://phabricator.fb.com/P20083553) ``` // $ du -h /home/tec/local/normal.4K.Snappy/db10077 // 6.1G /home/tec/local/normal.4K.Snappy/db10077 // $ du -h /home/tec/local/zero.8K.LZ4/db10077 // 6.4G /home/tec/local/zero.8K.LZ4/db10077 // Benchmarks for shard db10077 // _build/opt/rocks/benchmark/rocks_copy_benchmark \ // --normal_db_path="/home/tec/local/normal.4K.Snappy/db10077" \ // --zero_db_path="/home/tec/local/zero.8K.LZ4/db10077" // First run // ============================================================================ // rocks/benchmark/RocksCopyBenchmark.cpp relative time/iter iters/s // ============================================================================ // BM_StringCopy 1.73s 576.97m // BM_StringPiece 103.74% 1.67s 598.55m // ============================================================================ // Match rate : 1000000 / 1000000 // Second run // ============================================================================ // rocks/benchmark/RocksCopyBenchmark.cpp relative time/iter iters/s // ============================================================================ // BM_StringCopy 611.99ms 1.63 // BM_StringPiece 203.76% 300.35ms 3.33 // ============================================================================ // Match rate : 1000000 / 1000000 ``` Test Plan: Unit tests Reviewers: sdong, igor, anthony, yhchiang, rven Reviewed By: rven Subscribers: dhruba, lovro, adsharma Differential Revision: https://reviews.facebook.net/D48999
2015-12-16 20:08:30 +00:00
}
private:
// Clears heaps for both directions, used when changing direction or seeking
void ClearHeaps();
// Ensures that maxHeap_ is initialized when starting to go in the reverse
// direction
void InitMaxHeap();
bool is_arena_mode_;
const Comparator* comparator_;
autovector<IteratorWrapper, kNumIterReserve> children_;
// Cached pointer to child iterator with the current key, or nullptr if no
// child iterators are valid. This is the top of minHeap_ or maxHeap_
// depending on the direction.
IteratorWrapper* current_;
// Which direction is the iterator moving?
enum Direction {
kForward,
kReverse
};
Direction direction_;
MergerMinIterHeap minHeap_;
// Max heap is used for reverse iteration, which is way less common than
// forward. Lazily initialize it to save memory.
std::unique_ptr<MergerMaxIterHeap> maxHeap_;
PinnedIteratorsManager* pinned_iters_mgr_;
IteratorWrapper* CurrentForward() const {
assert(direction_ == kForward);
return !minHeap_.empty() ? minHeap_.top() : nullptr;
Replace std::priority_queue in MergingIterator with custom heap Summary: While profiling compaction in our service I noticed a lot of CPU (~15% of compaction) being spent in MergingIterator and key comparison. Looking at the code I found MergingIterator was (understandably) using std::priority_queue for the multiway merge. Keys in our dataset include sequence numbers that increase with time. Adjacent keys in an L0 file are very likely to be adjacent in the full database. Consequently, compaction will often pick a chunk of rows from the same L0 file before switching to another one. It would be great to avoid the O(log K) operation per row while compacting. This diff replaces std::priority_queue with a custom binary heap implementation. It has a "replace top" operation that is cheap when the new top is the same as the old one (i.e. the priority of the top entry is decreased but it still stays on top). Test Plan: make check To test the effect on performance, I generated databases with data patterns that mimic what I describe in the summary (rows have a mostly increasing sequence number). I see a 10-15% CPU decrease for compaction (and a matching throughput improvement on tmpfs). The exact improvement depends on the number of L0 files and the amount of locality. Performance on randomly distributed keys seems on par with the old code. Reviewers: kailiu, sdong, igor Reviewed By: igor Subscribers: yoshinorim, dhruba, tnovak Differential Revision: https://reviews.facebook.net/D29133
2015-07-06 11:24:09 +00:00
}
IteratorWrapper* CurrentReverse() const {
assert(direction_ == kReverse);
assert(maxHeap_);
return !maxHeap_->empty() ? maxHeap_->top() : nullptr;
Replace std::priority_queue in MergingIterator with custom heap Summary: While profiling compaction in our service I noticed a lot of CPU (~15% of compaction) being spent in MergingIterator and key comparison. Looking at the code I found MergingIterator was (understandably) using std::priority_queue for the multiway merge. Keys in our dataset include sequence numbers that increase with time. Adjacent keys in an L0 file are very likely to be adjacent in the full database. Consequently, compaction will often pick a chunk of rows from the same L0 file before switching to another one. It would be great to avoid the O(log K) operation per row while compacting. This diff replaces std::priority_queue with a custom binary heap implementation. It has a "replace top" operation that is cheap when the new top is the same as the old one (i.e. the priority of the top entry is decreased but it still stays on top). Test Plan: make check To test the effect on performance, I generated databases with data patterns that mimic what I describe in the summary (rows have a mostly increasing sequence number). I see a 10-15% CPU decrease for compaction (and a matching throughput improvement on tmpfs). The exact improvement depends on the number of L0 files and the amount of locality. Performance on randomly distributed keys seems on par with the old code. Reviewers: kailiu, sdong, igor Reviewed By: igor Subscribers: yoshinorim, dhruba, tnovak Differential Revision: https://reviews.facebook.net/D29133
2015-07-06 11:24:09 +00:00
}
};
void MergingIterator::ClearHeaps() {
minHeap_.clear();
if (maxHeap_) {
maxHeap_->clear();
}
}
void MergingIterator::InitMaxHeap() {
if (!maxHeap_) {
maxHeap_.reset(new MergerMaxIterHeap(comparator_));
}
}
InternalIterator* NewMergingIterator(const Comparator* cmp,
InternalIterator** list, int n,
Arena* arena) {
assert(n >= 0);
if (n == 0) {
return NewEmptyInternalIterator(arena);
} else if (n == 1) {
return list[0];
} else {
if (arena == nullptr) {
return new MergingIterator(cmp, list, n, false);
} else {
auto mem = arena->AllocateAligned(sizeof(MergingIterator));
return new (mem) MergingIterator(cmp, list, n, true);
}
}
}
MergeIteratorBuilder::MergeIteratorBuilder(const Comparator* comparator,
Arena* a)
: first_iter(nullptr), use_merging_iter(false), arena(a) {
auto mem = arena->AllocateAligned(sizeof(MergingIterator));
merge_iter = new (mem) MergingIterator(comparator, nullptr, 0, true);
}
void MergeIteratorBuilder::AddIterator(InternalIterator* iter) {
if (!use_merging_iter && first_iter != nullptr) {
merge_iter->AddIterator(first_iter);
use_merging_iter = true;
}
if (use_merging_iter) {
merge_iter->AddIterator(iter);
} else {
first_iter = iter;
}
}
InternalIterator* MergeIteratorBuilder::Finish() {
if (!use_merging_iter) {
return first_iter;
} else {
auto ret = merge_iter;
merge_iter = nullptr;
return ret;
}
}
} // namespace rocksdb