rocksdb/db/db_iter.cc

955 lines
31 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 "db/db_iter.h"
#include <stdexcept>
[RocksDB] [MergeOperator] The new Merge Interface! Uses merge sequences. Summary: Here are the major changes to the Merge Interface. It has been expanded to handle cases where the MergeOperator is not associative. It does so by stacking up merge operations while scanning through the key history (i.e.: during Get() or Compaction), until a valid Put/Delete/end-of-history is encountered; it then applies all of the merge operations in the correct sequence starting with the base/sentinel value. I have also introduced an "AssociativeMerge" function which allows the user to take advantage of associative merge operations (such as in the case of counters). The implementation will always attempt to merge the operations/operands themselves together when they are encountered, and will resort to the "stacking" method if and only if the "associative-merge" fails. This implementation is conjectured to allow MergeOperator to handle the general case, while still providing the user with the ability to take advantage of certain efficiencies in their own merge-operator / data-structure. NOTE: This is a preliminary diff. This must still go through a lot of review, revision, and testing. Feedback welcome! Test Plan: -This is a preliminary diff. I have only just begun testing/debugging it. -I will be testing this with the existing MergeOperator use-cases and unit-tests (counters, string-append, and redis-lists) -I will be "desk-checking" and walking through the code with the help gdb. -I will find a way of stress-testing the new interface / implementation using db_bench, db_test, merge_test, and/or db_stress. -I will ensure that my tests cover all cases: Get-Memtable, Get-Immutable-Memtable, Get-from-Disk, Iterator-Range-Scan, Flush-Memtable-to-L0, Compaction-L0-L1, Compaction-Ln-L(n+1), Put/Delete found, Put/Delete not-found, end-of-history, end-of-file, etc. -A lot of feedback from the reviewers. Reviewers: haobo, dhruba, zshao, emayanke Reviewed By: haobo CC: leveldb Differential Revision: https://reviews.facebook.net/D11499
2013-08-06 05:14:32 +02:00
#include <deque>
#include <string>
#include <limits>
#include "db/dbformat.h"
#include "db/filename.h"
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
#include "db/merge_context.h"
#include "port/port.h"
#include "rocksdb/env.h"
#include "rocksdb/iterator.h"
#include "rocksdb/merge_operator.h"
#include "rocksdb/options.h"
#include "table/internal_iterator.h"
#include "util/arena.h"
#include "util/logging.h"
#include "util/mutexlock.h"
#include "util/perf_context_imp.h"
#include "util/string_util.h"
namespace rocksdb {
#if 0
static void DumpInternalIter(Iterator* iter) {
for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
ParsedInternalKey k;
if (!ParseInternalKey(iter->key(), &k)) {
fprintf(stderr, "Corrupt '%s'\n", EscapeString(iter->key()).c_str());
} else {
fprintf(stderr, "@ '%s'\n", k.DebugString().c_str());
}
}
}
#endif
// Memtables and sstables that make the DB representation contain
// (userkey,seq,type) => uservalue entries. DBIter
// combines multiple entries for the same userkey found in the DB
// representation into a single entry while accounting for sequence
// numbers, deletion markers, overwrites, etc.
class DBIter: public Iterator {
public:
// The following is grossly complicated. TODO: clean it up
// Which direction is the iterator currently moving?
// (1) When moving forward, the internal iterator is positioned at
// the exact entry that yields this->key(), this->value()
// (2) When moving backwards, the internal iterator is positioned
// just before all entries whose user key == this->key().
enum Direction {
kForward,
kReverse
};
// LocalStatistics contain Statistics counters that will be aggregated per
// each iterator instance and then will be sent to the global statistics when
// the iterator is destroyed.
//
// The purpose of this approach is to avoid perf regression happening
// when multiple threads bump the atomic counters from a DBIter::Next().
struct LocalStatistics {
explicit LocalStatistics() { ResetCounters(); }
void ResetCounters() {
next_count_ = 0;
next_found_count_ = 0;
prev_count_ = 0;
prev_found_count_ = 0;
bytes_read_ = 0;
}
void BumpGlobalStatistics(Statistics* global_statistics) {
RecordTick(global_statistics, NUMBER_DB_NEXT, next_count_);
RecordTick(global_statistics, NUMBER_DB_NEXT_FOUND, next_found_count_);
RecordTick(global_statistics, NUMBER_DB_PREV, prev_count_);
RecordTick(global_statistics, NUMBER_DB_PREV_FOUND, prev_found_count_);
RecordTick(global_statistics, ITER_BYTES_READ, bytes_read_);
ResetCounters();
}
// Map to Tickers::NUMBER_DB_NEXT
uint64_t next_count_;
// Map to Tickers::NUMBER_DB_NEXT_FOUND
uint64_t next_found_count_;
// Map to Tickers::NUMBER_DB_PREV
uint64_t prev_count_;
// Map to Tickers::NUMBER_DB_PREV_FOUND
uint64_t prev_found_count_;
// Map to Tickers::ITER_BYTES_READ
uint64_t bytes_read_;
};
DBIter(Env* env, const ImmutableCFOptions& ioptions, const Comparator* cmp,
InternalIterator* iter, SequenceNumber s, bool arena_mode,
uint64_t max_sequential_skip_in_iterations, uint64_t version_number,
const Slice* iterate_upper_bound = nullptr,
bool prefix_same_as_start = false)
: arena_mode_(arena_mode),
env_(env),
logger_(ioptions.info_log),
user_comparator_(cmp),
user_merge_operator_(ioptions.merge_operator),
iter_(iter),
sequence_(s),
direction_(kForward),
valid_(false),
current_entry_is_merged_(false),
statistics_(ioptions.statistics),
version_number_(version_number),
iterate_upper_bound_(iterate_upper_bound),
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 21:08:30 +01:00
prefix_same_as_start_(prefix_same_as_start),
iter_pinned_(false) {
RecordTick(statistics_, NO_ITERATORS);
prefix_extractor_ = ioptions.prefix_extractor;
max_skip_ = max_sequential_skip_in_iterations;
}
virtual ~DBIter() {
RecordTick(statistics_, NO_ITERATORS, -1);
local_stats_.BumpGlobalStatistics(statistics_);
if (!arena_mode_) {
delete iter_;
} else {
iter_->~InternalIterator();
}
}
virtual void SetIter(InternalIterator* iter) {
assert(iter_ == nullptr);
iter_ = iter;
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 21:08:30 +01:00
if (iter_ && iter_pinned_) {
iter_->PinData();
}
}
virtual bool Valid() const override { return valid_; }
virtual Slice key() const override {
assert(valid_);
return saved_key_.GetKey();
}
virtual Slice value() const override {
assert(valid_);
return (direction_ == kForward && !current_entry_is_merged_) ?
iter_->value() : saved_value_;
}
virtual Status status() const override {
if (status_.ok()) {
return iter_->status();
} else {
return status_;
}
}
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 21:08:30 +01:00
virtual Status PinData() {
Status s;
if (iter_) {
s = iter_->PinData();
}
if (s.ok()) {
// Even if iter_ is nullptr, we set iter_pinned_ to true so that when
// iter_ is updated using SetIter, we Pin it.
iter_pinned_ = true;
}
return s;
}
virtual Status ReleasePinnedData() {
Status s;
if (iter_) {
s = iter_->ReleasePinnedData();
}
if (s.ok()) {
iter_pinned_ = false;
}
return s;
}
virtual Status GetProperty(std::string prop_name,
std::string* prop) override {
if (prop == nullptr) {
return Status::InvalidArgument("prop is nullptr");
}
if (prop_name == "rocksdb.iterator.super-version-number") {
// First try to pass the value returned from inner iterator.
if (!iter_->GetProperty(prop_name, prop).ok()) {
*prop = ToString(version_number_);
}
return Status::OK();
} else if (prop_name == "rocksdb.iterator.is-key-pinned") {
if (valid_) {
*prop = (iter_pinned_ && saved_key_.IsKeyPinned()) ? "1" : "0";
} else {
*prop = "Iterator is not valid.";
}
return Status::OK();
}
return Status::InvalidArgument("Undentified property.");
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 21:08:30 +01:00
}
virtual void Next() override;
virtual void Prev() override;
virtual void Seek(const Slice& target) override;
virtual void SeekToFirst() override;
virtual void SeekToLast() override;
private:
void ReverseToBackward();
void PrevInternal();
void FindParseableKey(ParsedInternalKey* ikey, Direction direction);
bool FindValueForCurrentKey();
bool FindValueForCurrentKeyUsingSeek();
void FindPrevUserKey();
void FindNextUserKey();
inline void FindNextUserEntry(bool skipping);
void FindNextUserEntryInternal(bool skipping);
bool ParseKey(ParsedInternalKey* key);
void MergeValuesNewToOld();
inline void ClearSavedValue() {
if (saved_value_.capacity() > 1048576) {
std::string empty;
swap(empty, saved_value_);
} else {
saved_value_.clear();
}
}
const SliceTransform* prefix_extractor_;
bool arena_mode_;
Env* const env_;
Logger* logger_;
const Comparator* const user_comparator_;
const MergeOperator* const user_merge_operator_;
InternalIterator* iter_;
SequenceNumber const sequence_;
Status status_;
IterKey saved_key_;
std::string saved_value_;
Direction direction_;
bool valid_;
bool current_entry_is_merged_;
Statistics* statistics_;
uint64_t max_skip_;
uint64_t version_number_;
const Slice* iterate_upper_bound_;
IterKey prefix_start_;
bool prefix_same_as_start_;
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 21:08:30 +01:00
bool iter_pinned_;
2016-01-07 16:59:14 +01:00
// List of operands for merge operator.
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
MergeContext merge_context_;
LocalStatistics local_stats_;
// No copying allowed
DBIter(const DBIter&);
void operator=(const DBIter&);
};
inline bool DBIter::ParseKey(ParsedInternalKey* ikey) {
if (!ParseInternalKey(iter_->key(), ikey)) {
status_ = Status::Corruption("corrupted internal key in DBIter");
Log(InfoLogLevel::ERROR_LEVEL,
logger_, "corrupted internal key in DBIter: %s",
iter_->key().ToString(true).c_str());
return false;
} else {
return true;
}
}
void DBIter::Next() {
assert(valid_);
if (direction_ == kReverse) {
FindNextUserKey();
direction_ = kForward;
if (!iter_->Valid()) {
iter_->SeekToFirst();
}
} else if (iter_->Valid() && !current_entry_is_merged_) {
// If the current value is not a merge, the iter position is the
// current key, which is already returned. We can safely issue a
// Next() without checking the current key.
// If the current key is a merge, very likely iter already points
// to the next internal position.
iter_->Next();
PERF_COUNTER_ADD(internal_key_skipped_count, 1);
}
if (statistics_ != nullptr) {
local_stats_.next_count_++;
}
// Now we point to the next internal position, for both of merge and
// not merge cases.
if (!iter_->Valid()) {
valid_ = false;
return;
}
FindNextUserEntry(true /* skipping the current user key */);
if (valid_ && prefix_extractor_ && prefix_same_as_start_ &&
prefix_extractor_->Transform(saved_key_.GetKey())
.compare(prefix_start_.GetKey()) != 0) {
valid_ = false;
}
if (statistics_ != nullptr && valid_) {
local_stats_.next_found_count_++;
local_stats_.bytes_read_ += (key().size() + value().size());
}
}
// PRE: saved_key_ has the current user key if skipping
// POST: saved_key_ should have the next user key if valid_,
// if the current entry is a result of merge
// current_entry_is_merged_ => true
// saved_value_ => the merged value
//
// NOTE: In between, saved_key_ can point to a user key that has
// a delete marker
inline void DBIter::FindNextUserEntry(bool skipping) {
PERF_TIMER_GUARD(find_next_user_entry_time);
FindNextUserEntryInternal(skipping);
}
// Actual implementation of DBIter::FindNextUserEntry()
void DBIter::FindNextUserEntryInternal(bool skipping) {
// Loop until we hit an acceptable entry to yield
assert(iter_->Valid());
assert(direction_ == kForward);
current_entry_is_merged_ = false;
uint64_t num_skipped = 0;
do {
ParsedInternalKey ikey;
if (ParseKey(&ikey)) {
if (iterate_upper_bound_ != nullptr &&
user_comparator_->Compare(ikey.user_key, *iterate_upper_bound_) >= 0) {
break;
}
if (ikey.sequence <= sequence_) {
if (skipping &&
user_comparator_->Compare(ikey.user_key, saved_key_.GetKey()) <= 0) {
num_skipped++; // skip this entry
PERF_COUNTER_ADD(internal_key_skipped_count, 1);
} else {
switch (ikey.type) {
case kTypeDeletion:
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
case kTypeSingleDeletion:
// Arrange to skip all upcoming entries for this key since
// they are hidden by this deletion.
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 21:08:30 +01:00
saved_key_.SetKey(ikey.user_key,
!iter_->IsKeyPinned() /* copy */);
skipping = true;
num_skipped = 0;
PERF_COUNTER_ADD(internal_delete_skipped_count, 1);
break;
case kTypeValue:
valid_ = true;
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 21:08:30 +01:00
saved_key_.SetKey(ikey.user_key,
!iter_->IsKeyPinned() /* copy */);
return;
case kTypeMerge:
// By now, we are sure the current ikey is going to yield a value
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 21:08:30 +01:00
saved_key_.SetKey(ikey.user_key,
!iter_->IsKeyPinned() /* copy */);
current_entry_is_merged_ = true;
valid_ = true;
MergeValuesNewToOld(); // Go to a different state machine
return;
default:
assert(false);
break;
}
}
}
}
// If we have sequentially iterated via numerous keys and still not
// found the next user-key, then it is better to seek so that we can
2015-04-25 11:14:27 +02:00
// avoid too many key comparisons. We seek to the last occurrence of
// our current key by looking for sequence number 0 and type deletion
// (the smallest type).
if (skipping && num_skipped > max_skip_) {
num_skipped = 0;
std::string last_key;
AppendInternalKey(&last_key, ParsedInternalKey(saved_key_.GetKey(), 0,
kTypeDeletion));
iter_->Seek(last_key);
RecordTick(statistics_, NUMBER_OF_RESEEKS_IN_ITERATION);
} else {
iter_->Next();
}
} while (iter_->Valid());
valid_ = false;
}
// Merge values of the same user key starting from the current iter_ position
// Scan from the newer entries to older entries.
// PRE: iter_->key() points to the first merge type entry
// saved_key_ stores the user key
// POST: saved_value_ has the merged value for the user key
// iter_ points to the next entry (or invalid)
void DBIter::MergeValuesNewToOld() {
if (!user_merge_operator_) {
Log(InfoLogLevel::ERROR_LEVEL,
logger_, "Options::merge_operator is null.");
status_ = Status::InvalidArgument("user_merge_operator_ must be set.");
valid_ = false;
return;
}
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.Clear();
[RocksDB] [MergeOperator] The new Merge Interface! Uses merge sequences. Summary: Here are the major changes to the Merge Interface. It has been expanded to handle cases where the MergeOperator is not associative. It does so by stacking up merge operations while scanning through the key history (i.e.: during Get() or Compaction), until a valid Put/Delete/end-of-history is encountered; it then applies all of the merge operations in the correct sequence starting with the base/sentinel value. I have also introduced an "AssociativeMerge" function which allows the user to take advantage of associative merge operations (such as in the case of counters). The implementation will always attempt to merge the operations/operands themselves together when they are encountered, and will resort to the "stacking" method if and only if the "associative-merge" fails. This implementation is conjectured to allow MergeOperator to handle the general case, while still providing the user with the ability to take advantage of certain efficiencies in their own merge-operator / data-structure. NOTE: This is a preliminary diff. This must still go through a lot of review, revision, and testing. Feedback welcome! Test Plan: -This is a preliminary diff. I have only just begun testing/debugging it. -I will be testing this with the existing MergeOperator use-cases and unit-tests (counters, string-append, and redis-lists) -I will be "desk-checking" and walking through the code with the help gdb. -I will find a way of stress-testing the new interface / implementation using db_bench, db_test, merge_test, and/or db_stress. -I will ensure that my tests cover all cases: Get-Memtable, Get-Immutable-Memtable, Get-from-Disk, Iterator-Range-Scan, Flush-Memtable-to-L0, Compaction-L0-L1, Compaction-Ln-L(n+1), Put/Delete found, Put/Delete not-found, end-of-history, end-of-file, etc. -A lot of feedback from the reviewers. Reviewers: haobo, dhruba, zshao, emayanke Reviewed By: haobo CC: leveldb Differential Revision: https://reviews.facebook.net/D11499
2013-08-06 05:14:32 +02:00
// Start the merge process by pushing the first operand
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.PushOperand(iter_->value());
ParsedInternalKey ikey;
for (iter_->Next(); iter_->Valid(); iter_->Next()) {
if (!ParseKey(&ikey)) {
// skip corrupted key
continue;
}
if (!user_comparator_->Equal(ikey.user_key, saved_key_.GetKey())) {
// hit the next user key, stop right here
break;
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
} else if (kTypeDeletion == ikey.type || kTypeSingleDeletion == ikey.type) {
// hit a delete with the same user key, stop right here
// iter_ is positioned after delete
iter_->Next();
break;
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
} else if (kTypeValue == ikey.type) {
[RocksDB] [MergeOperator] The new Merge Interface! Uses merge sequences. Summary: Here are the major changes to the Merge Interface. It has been expanded to handle cases where the MergeOperator is not associative. It does so by stacking up merge operations while scanning through the key history (i.e.: during Get() or Compaction), until a valid Put/Delete/end-of-history is encountered; it then applies all of the merge operations in the correct sequence starting with the base/sentinel value. I have also introduced an "AssociativeMerge" function which allows the user to take advantage of associative merge operations (such as in the case of counters). The implementation will always attempt to merge the operations/operands themselves together when they are encountered, and will resort to the "stacking" method if and only if the "associative-merge" fails. This implementation is conjectured to allow MergeOperator to handle the general case, while still providing the user with the ability to take advantage of certain efficiencies in their own merge-operator / data-structure. NOTE: This is a preliminary diff. This must still go through a lot of review, revision, and testing. Feedback welcome! Test Plan: -This is a preliminary diff. I have only just begun testing/debugging it. -I will be testing this with the existing MergeOperator use-cases and unit-tests (counters, string-append, and redis-lists) -I will be "desk-checking" and walking through the code with the help gdb. -I will find a way of stress-testing the new interface / implementation using db_bench, db_test, merge_test, and/or db_stress. -I will ensure that my tests cover all cases: Get-Memtable, Get-Immutable-Memtable, Get-from-Disk, Iterator-Range-Scan, Flush-Memtable-to-L0, Compaction-L0-L1, Compaction-Ln-L(n+1), Put/Delete found, Put/Delete not-found, end-of-history, end-of-file, etc. -A lot of feedback from the reviewers. Reviewers: haobo, dhruba, zshao, emayanke Reviewed By: haobo CC: leveldb Differential Revision: https://reviews.facebook.net/D11499
2013-08-06 05:14:32 +02:00
// hit a put, merge the put value with operands and store the
// final result in saved_value_. We are done!
// ignore corruption if there is any.
const Slice val = iter_->value();
{
StopWatchNano timer(env_, statistics_ != nullptr);
PERF_TIMER_GUARD(merge_operator_time_nanos);
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
user_merge_operator_->FullMerge(ikey.user_key, &val,
merge_context_.GetOperands(),
&saved_value_, logger_);
RecordTick(statistics_, MERGE_OPERATION_TOTAL_TIME,
timer.ElapsedNanos());
}
// iter_ is positioned after put
iter_->Next();
return;
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
} else if (kTypeMerge == ikey.type) {
[RocksDB] [MergeOperator] The new Merge Interface! Uses merge sequences. Summary: Here are the major changes to the Merge Interface. It has been expanded to handle cases where the MergeOperator is not associative. It does so by stacking up merge operations while scanning through the key history (i.e.: during Get() or Compaction), until a valid Put/Delete/end-of-history is encountered; it then applies all of the merge operations in the correct sequence starting with the base/sentinel value. I have also introduced an "AssociativeMerge" function which allows the user to take advantage of associative merge operations (such as in the case of counters). The implementation will always attempt to merge the operations/operands themselves together when they are encountered, and will resort to the "stacking" method if and only if the "associative-merge" fails. This implementation is conjectured to allow MergeOperator to handle the general case, while still providing the user with the ability to take advantage of certain efficiencies in their own merge-operator / data-structure. NOTE: This is a preliminary diff. This must still go through a lot of review, revision, and testing. Feedback welcome! Test Plan: -This is a preliminary diff. I have only just begun testing/debugging it. -I will be testing this with the existing MergeOperator use-cases and unit-tests (counters, string-append, and redis-lists) -I will be "desk-checking" and walking through the code with the help gdb. -I will find a way of stress-testing the new interface / implementation using db_bench, db_test, merge_test, and/or db_stress. -I will ensure that my tests cover all cases: Get-Memtable, Get-Immutable-Memtable, Get-from-Disk, Iterator-Range-Scan, Flush-Memtable-to-L0, Compaction-L0-L1, Compaction-Ln-L(n+1), Put/Delete found, Put/Delete not-found, end-of-history, end-of-file, etc. -A lot of feedback from the reviewers. Reviewers: haobo, dhruba, zshao, emayanke Reviewed By: haobo CC: leveldb Differential Revision: https://reviews.facebook.net/D11499
2013-08-06 05:14:32 +02:00
// hit a merge, add the value as an operand and run associative merge.
// when complete, add result to operands and continue.
const Slice& val = iter_->value();
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.PushOperand(val);
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
} else {
assert(false);
}
}
{
StopWatchNano timer(env_, statistics_ != nullptr);
PERF_TIMER_GUARD(merge_operator_time_nanos);
// we either exhausted all internal keys under this user key, or hit
// a deletion marker.
// feed null as the existing value to the merge operator, such that
// client can differentiate this scenario and do things accordingly.
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
user_merge_operator_->FullMerge(saved_key_.GetKey(), nullptr,
merge_context_.GetOperands(), &saved_value_,
logger_);
RecordTick(statistics_, MERGE_OPERATION_TOTAL_TIME, timer.ElapsedNanos());
}
}
void DBIter::Prev() {
assert(valid_);
if (direction_ == kForward) {
ReverseToBackward();
}
PrevInternal();
if (statistics_ != nullptr) {
local_stats_.prev_count_++;
if (valid_) {
local_stats_.prev_found_count_++;
local_stats_.bytes_read_ += (key().size() + value().size());
}
}
if (valid_ && prefix_extractor_ && prefix_same_as_start_ &&
prefix_extractor_->Transform(saved_key_.GetKey())
.compare(prefix_start_.GetKey()) != 0) {
valid_ = false;
}
}
void DBIter::ReverseToBackward() {
if (current_entry_is_merged_) {
// Not placed in the same key. Need to call Prev() until finding the
// previous key.
if (!iter_->Valid()) {
iter_->SeekToLast();
}
ParsedInternalKey ikey;
FindParseableKey(&ikey, kReverse);
while (iter_->Valid() &&
user_comparator_->Compare(ikey.user_key, saved_key_.GetKey()) > 0) {
iter_->Prev();
FindParseableKey(&ikey, kReverse);
}
}
#ifndef NDEBUG
if (iter_->Valid()) {
ParsedInternalKey ikey;
assert(ParseKey(&ikey));
assert(user_comparator_->Compare(ikey.user_key, saved_key_.GetKey()) <= 0);
}
#endif
FindPrevUserKey();
direction_ = kReverse;
}
void DBIter::PrevInternal() {
if (!iter_->Valid()) {
valid_ = false;
return;
}
ParsedInternalKey ikey;
while (iter_->Valid()) {
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 21:08:30 +01:00
saved_key_.SetKey(ExtractUserKey(iter_->key()),
!iter_->IsKeyPinned() /* copy */);
if (FindValueForCurrentKey()) {
valid_ = true;
if (!iter_->Valid()) {
return;
}
FindParseableKey(&ikey, kReverse);
if (user_comparator_->Equal(ikey.user_key, saved_key_.GetKey())) {
FindPrevUserKey();
}
return;
}
if (!iter_->Valid()) {
break;
}
FindParseableKey(&ikey, kReverse);
if (user_comparator_->Equal(ikey.user_key, saved_key_.GetKey())) {
FindPrevUserKey();
}
}
// We haven't found any key - iterator is not valid
assert(!iter_->Valid());
valid_ = false;
}
// This function checks, if the entry with biggest sequence_number <= sequence_
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
// is non kTypeDeletion or kTypeSingleDeletion. If it's not, we save value in
// saved_value_
bool DBIter::FindValueForCurrentKey() {
assert(iter_->Valid());
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.Clear();
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
// last entry before merge (could be kTypeDeletion, kTypeSingleDeletion or
// kTypeValue)
ValueType last_not_merge_type = kTypeDeletion;
ValueType last_key_entry_type = kTypeDeletion;
ParsedInternalKey ikey;
FindParseableKey(&ikey, kReverse);
size_t num_skipped = 0;
while (iter_->Valid() && ikey.sequence <= sequence_ &&
user_comparator_->Equal(ikey.user_key, saved_key_.GetKey())) {
// We iterate too much: let's use Seek() to avoid too much key comparisons
if (num_skipped >= max_skip_) {
return FindValueForCurrentKeyUsingSeek();
}
last_key_entry_type = ikey.type;
switch (last_key_entry_type) {
case kTypeValue:
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.Clear();
saved_value_ = iter_->value().ToString();
last_not_merge_type = kTypeValue;
break;
case kTypeDeletion:
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
case kTypeSingleDeletion:
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.Clear();
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
last_not_merge_type = last_key_entry_type;
PERF_COUNTER_ADD(internal_delete_skipped_count, 1);
break;
case kTypeMerge:
assert(user_merge_operator_ != nullptr);
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.PushOperandBack(iter_->value());
break;
default:
assert(false);
}
PERF_COUNTER_ADD(internal_key_skipped_count, 1);
assert(user_comparator_->Equal(ikey.user_key, saved_key_.GetKey()));
iter_->Prev();
++num_skipped;
FindParseableKey(&ikey, kReverse);
}
switch (last_key_entry_type) {
case kTypeDeletion:
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
case kTypeSingleDeletion:
valid_ = false;
return false;
case kTypeMerge:
if (last_not_merge_type == kTypeDeletion) {
StopWatchNano timer(env_, statistics_ != nullptr);
PERF_TIMER_GUARD(merge_operator_time_nanos);
2016-01-07 16:59:14 +01:00
user_merge_operator_->FullMerge(saved_key_.GetKey(), nullptr,
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.GetOperands(),
&saved_value_, logger_);
RecordTick(statistics_, MERGE_OPERATION_TOTAL_TIME,
timer.ElapsedNanos());
} else {
assert(last_not_merge_type == kTypeValue);
std::string last_put_value = saved_value_;
Slice temp_slice(last_put_value);
{
StopWatchNano timer(env_, statistics_ != nullptr);
PERF_TIMER_GUARD(merge_operator_time_nanos);
user_merge_operator_->FullMerge(saved_key_.GetKey(), &temp_slice,
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.GetOperands(),
&saved_value_, logger_);
RecordTick(statistics_, MERGE_OPERATION_TOTAL_TIME,
timer.ElapsedNanos());
}
}
break;
case kTypeValue:
// do nothing - we've already has value in saved_value_
break;
default:
assert(false);
break;
}
valid_ = true;
return true;
}
// This function is used in FindValueForCurrentKey.
// We use Seek() function instead of Prev() to find necessary value
bool DBIter::FindValueForCurrentKeyUsingSeek() {
std::string last_key;
AppendInternalKey(&last_key, ParsedInternalKey(saved_key_.GetKey(), sequence_,
kValueTypeForSeek));
iter_->Seek(last_key);
RecordTick(statistics_, NUMBER_OF_RESEEKS_IN_ITERATION);
// assume there is at least one parseable key for this user key
ParsedInternalKey ikey;
FindParseableKey(&ikey, kForward);
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
if (ikey.type == kTypeValue || ikey.type == kTypeDeletion ||
ikey.type == kTypeSingleDeletion) {
if (ikey.type == kTypeValue) {
saved_value_ = iter_->value().ToString();
valid_ = true;
return true;
}
valid_ = false;
return false;
}
// kTypeMerge. We need to collect all kTypeMerge values and save them
// in operands
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.Clear();
while (iter_->Valid() &&
user_comparator_->Equal(ikey.user_key, saved_key_.GetKey()) &&
ikey.type == kTypeMerge) {
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
merge_context_.PushOperand(iter_->value());
iter_->Next();
FindParseableKey(&ikey, kForward);
}
if (!iter_->Valid() ||
!user_comparator_->Equal(ikey.user_key, saved_key_.GetKey()) ||
Support for SingleDelete() Summary: This patch fixes #7460559. It introduces SingleDelete as a new database operation. This operation can be used to delete keys that were never overwritten (no put following another put of the same key). If an overwritten key is single deleted the behavior is undefined. Single deletion of a non-existent key has no effect but multiple consecutive single deletions are not allowed (see limitations). In contrast to the conventional Delete() operation, the deletion entry is removed along with the value when the two are lined up in a compaction. Note: The semantics are similar to @igor's prototype that allowed to have this behavior on the granularity of a column family ( https://reviews.facebook.net/D42093 ). This new patch, however, is more aggressive when it comes to removing tombstones: It removes the SingleDelete together with the value whenever there is no snapshot between them while the older patch only did this when the sequence number of the deletion was older than the earliest snapshot. Most of the complex additions are in the Compaction Iterator, all other changes should be relatively straightforward. The patch also includes basic support for single deletions in db_stress and db_bench. Limitations: - Not compatible with cuckoo hash tables - Single deletions cannot be used in combination with merges and normal deletions on the same key (other keys are not affected by this) - Consecutive single deletions are currently not allowed (and older version of this patch supported this so it could be resurrected if needed) Test Plan: make all check Reviewers: yhchiang, sdong, rven, anthony, yoshinorim, igor Reviewed By: igor Subscribers: maykov, dhruba, leveldb Differential Revision: https://reviews.facebook.net/D43179
2015-09-17 20:42:56 +02:00
ikey.type == kTypeDeletion || ikey.type == kTypeSingleDeletion) {
{
StopWatchNano timer(env_, statistics_ != nullptr);
PERF_TIMER_GUARD(merge_operator_time_nanos);
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
user_merge_operator_->FullMerge(saved_key_.GetKey(), nullptr,
merge_context_.GetOperands(),
&saved_value_, logger_);
RecordTick(statistics_, MERGE_OPERATION_TOTAL_TIME, timer.ElapsedNanos());
}
// Make iter_ valid and point to saved_key_
if (!iter_->Valid() ||
!user_comparator_->Equal(ikey.user_key, saved_key_.GetKey())) {
iter_->Seek(last_key);
RecordTick(statistics_, NUMBER_OF_RESEEKS_IN_ITERATION);
}
valid_ = true;
return true;
}
const Slice& val = iter_->value();
{
StopWatchNano timer(env_, statistics_ != nullptr);
PERF_TIMER_GUARD(merge_operator_time_nanos);
Eliminate std::deque initialization while iterating over merge operands Summary: This patch is similar to D52563, When we iterate over a DB with merge operands we keep creating std::queue to store the operands, optimize this by reusing merge_operands_ data member Before the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.757 micros/op 266141 ops/sec; 29.4 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.413 micros/op 2423538 ops/sec; 268.1 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.451 micros/op 2219071 ops/sec; 245.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.420 micros/op 2382039 ops/sec; 263.5 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.408 micros/op 2452017 ops/sec; 271.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.947 micros/op 253376 ops/sec; 28.0 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.441 micros/op 2266473 ops/sec; 250.7 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.471 micros/op 2122033 ops/sec; 234.8 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.440 micros/op 2271407 ops/sec; 251.3 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.429 micros/op 2331471 ops/sec; 257.9 MB/s ``` with the patch ``` ./db_bench --benchmarks="mergerandom,readseq,readseq,readseq,readseq" --db="/dev/shm/bench_merge_memcpy_on_the_fly/" --merge_operator="put" --merge_keys=10000 --num=10000 DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 4.080 micros/op 245092 ops/sec; 27.1 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.308 micros/op 3241843 ops/sec; 358.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.312 micros/op 3200408 ops/sec; 354.0 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.332 micros/op 3013962 ops/sec; 333.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.300 micros/op 3328017 ops/sec; 368.2 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] mergerandom : 3.973 micros/op 251705 ops/sec; 27.8 MB/s ( updates:10000) DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.320 micros/op 3123752 ops/sec; 345.6 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.335 micros/op 2986641 ops/sec; 330.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.339 micros/op 2950047 ops/sec; 326.4 MB/s DB path: [/dev/shm/bench_merge_memcpy_on_the_fly/] readseq : 0.319 micros/op 3131565 ops/sec; 346.4 MB/s ``` Test Plan: make check -j64 Reviewers: yhchiang, andrewkr, sdong Reviewed By: sdong Subscribers: andrewkr, dhruba Differential Revision: https://reviews.facebook.net/D56031
2016-04-02 00:48:55 +02:00
user_merge_operator_->FullMerge(saved_key_.GetKey(), &val,
merge_context_.GetOperands(), &saved_value_,
logger_);
RecordTick(statistics_, MERGE_OPERATION_TOTAL_TIME, timer.ElapsedNanos());
}
valid_ = true;
return true;
}
// Used in Next to change directions
// Go to next user key
// Don't use Seek(),
// because next user key will be very close
void DBIter::FindNextUserKey() {
if (!iter_->Valid()) {
return;
}
ParsedInternalKey ikey;
FindParseableKey(&ikey, kForward);
while (iter_->Valid() &&
!user_comparator_->Equal(ikey.user_key, saved_key_.GetKey())) {
iter_->Next();
FindParseableKey(&ikey, kForward);
}
}
// Go to previous user_key
void DBIter::FindPrevUserKey() {
if (!iter_->Valid()) {
return;
}
size_t num_skipped = 0;
ParsedInternalKey ikey;
FindParseableKey(&ikey, kReverse);
int cmp;
while (iter_->Valid() && ((cmp = user_comparator_->Compare(
ikey.user_key, saved_key_.GetKey())) == 0 ||
(cmp > 0 && ikey.sequence > sequence_))) {
if (cmp == 0) {
if (num_skipped >= max_skip_) {
num_skipped = 0;
IterKey last_key;
last_key.SetInternalKey(ParsedInternalKey(
saved_key_.GetKey(), kMaxSequenceNumber, kValueTypeForSeek));
iter_->Seek(last_key.GetKey());
RecordTick(statistics_, NUMBER_OF_RESEEKS_IN_ITERATION);
} else {
++num_skipped;
}
}
iter_->Prev();
FindParseableKey(&ikey, kReverse);
}
}
// Skip all unparseable keys
void DBIter::FindParseableKey(ParsedInternalKey* ikey, Direction direction) {
while (iter_->Valid() && !ParseKey(ikey)) {
if (direction == kReverse) {
iter_->Prev();
} else {
iter_->Next();
}
}
}
void DBIter::Seek(const Slice& target) {
StopWatch sw(env_, statistics_, DB_SEEK);
saved_key_.Clear();
// now savved_key is used to store internal key.
saved_key_.SetInternalKey(target, sequence_);
{
PERF_TIMER_GUARD(seek_internal_seek_time);
iter_->Seek(saved_key_.GetKey());
}
RecordTick(statistics_, NUMBER_DB_SEEK);
if (iter_->Valid()) {
direction_ = kForward;
ClearSavedValue();
FindNextUserEntry(false /* not skipping */);
if (statistics_ != nullptr) {
if (valid_) {
RecordTick(statistics_, NUMBER_DB_SEEK_FOUND);
RecordTick(statistics_, ITER_BYTES_READ, key().size() + value().size());
}
}
} else {
valid_ = false;
}
if (valid_ && prefix_extractor_ && prefix_same_as_start_) {
prefix_start_.SetKey(prefix_extractor_->Transform(target));
}
}
void DBIter::SeekToFirst() {
// Don't use iter_::Seek() if we set a prefix extractor
// because prefix seek will be used.
if (prefix_extractor_ != nullptr) {
max_skip_ = std::numeric_limits<uint64_t>::max();
}
direction_ = kForward;
ClearSavedValue();
{
PERF_TIMER_GUARD(seek_internal_seek_time);
iter_->SeekToFirst();
}
RecordTick(statistics_, NUMBER_DB_SEEK);
if (iter_->Valid()) {
FindNextUserEntry(false /* not skipping */);
if (statistics_ != nullptr) {
if (valid_) {
RecordTick(statistics_, NUMBER_DB_SEEK_FOUND);
RecordTick(statistics_, ITER_BYTES_READ, key().size() + value().size());
}
}
} else {
valid_ = false;
}
if (valid_ && prefix_extractor_ && prefix_same_as_start_) {
prefix_start_.SetKey(prefix_extractor_->Transform(saved_key_.GetKey()));
}
}
void DBIter::SeekToLast() {
// Don't use iter_::Seek() if we set a prefix extractor
// because prefix seek will be used.
if (prefix_extractor_ != nullptr) {
max_skip_ = std::numeric_limits<uint64_t>::max();
}
direction_ = kReverse;
ClearSavedValue();
{
PERF_TIMER_GUARD(seek_internal_seek_time);
iter_->SeekToLast();
}
// When the iterate_upper_bound is set to a value,
// it will seek to the last key before the
// ReadOptions.iterate_upper_bound
if (iter_->Valid() && iterate_upper_bound_ != nullptr) {
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 21:08:30 +01:00
saved_key_.SetKey(*iterate_upper_bound_, false /* copy */);
std::string last_key;
AppendInternalKey(&last_key,
ParsedInternalKey(saved_key_.GetKey(), kMaxSequenceNumber,
kValueTypeForSeek));
iter_->Seek(last_key);
if (!iter_->Valid()) {
iter_->SeekToLast();
} else {
iter_->Prev();
if (!iter_->Valid()) {
valid_ = false;
return;
}
}
}
PrevInternal();
if (statistics_ != nullptr) {
RecordTick(statistics_, NUMBER_DB_SEEK);
if (valid_) {
RecordTick(statistics_, NUMBER_DB_SEEK_FOUND);
RecordTick(statistics_, ITER_BYTES_READ, key().size() + value().size());
}
}
if (valid_ && prefix_extractor_ && prefix_same_as_start_) {
prefix_start_.SetKey(prefix_extractor_->Transform(saved_key_.GetKey()));
}
}
Iterator* NewDBIterator(Env* env, const ImmutableCFOptions& ioptions,
const Comparator* user_key_comparator,
InternalIterator* internal_iter,
const SequenceNumber& sequence,
uint64_t max_sequential_skip_in_iterations,
uint64_t version_number,
const Slice* iterate_upper_bound,
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 21:08:30 +01:00
bool prefix_same_as_start, bool pin_data) {
DBIter* db_iter =
new DBIter(env, ioptions, user_key_comparator, internal_iter, sequence,
false, max_sequential_skip_in_iterations, version_number,
iterate_upper_bound, prefix_same_as_start);
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 21:08:30 +01:00
if (pin_data) {
db_iter->PinData();
}
return db_iter;
}
ArenaWrappedDBIter::~ArenaWrappedDBIter() { db_iter_->~DBIter(); }
void ArenaWrappedDBIter::SetDBIter(DBIter* iter) { db_iter_ = iter; }
void ArenaWrappedDBIter::SetIterUnderDBIter(InternalIterator* iter) {
static_cast<DBIter*>(db_iter_)->SetIter(iter);
}
inline bool ArenaWrappedDBIter::Valid() const { return db_iter_->Valid(); }
inline void ArenaWrappedDBIter::SeekToFirst() { db_iter_->SeekToFirst(); }
inline void ArenaWrappedDBIter::SeekToLast() { db_iter_->SeekToLast(); }
inline void ArenaWrappedDBIter::Seek(const Slice& target) {
db_iter_->Seek(target);
}
inline void ArenaWrappedDBIter::Next() { db_iter_->Next(); }
inline void ArenaWrappedDBIter::Prev() { db_iter_->Prev(); }
inline Slice ArenaWrappedDBIter::key() const { return db_iter_->key(); }
inline Slice ArenaWrappedDBIter::value() const { return db_iter_->value(); }
inline Status ArenaWrappedDBIter::status() const { return db_iter_->status(); }
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 21:08:30 +01:00
inline Status ArenaWrappedDBIter::PinData() { return db_iter_->PinData(); }
inline Status ArenaWrappedDBIter::GetProperty(std::string prop_name,
std::string* prop) {
return db_iter_->GetProperty(prop_name, prop);
}
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 21:08:30 +01:00
inline Status ArenaWrappedDBIter::ReleasePinnedData() {
return db_iter_->ReleasePinnedData();
}
void ArenaWrappedDBIter::RegisterCleanup(CleanupFunction function, void* arg1,
void* arg2) {
db_iter_->RegisterCleanup(function, arg1, arg2);
}
ArenaWrappedDBIter* NewArenaWrappedDbIterator(
Env* env, const ImmutableCFOptions& ioptions,
const Comparator* user_key_comparator, const SequenceNumber& sequence,
uint64_t max_sequential_skip_in_iterations, uint64_t version_number,
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 21:08:30 +01:00
const Slice* iterate_upper_bound, bool prefix_same_as_start,
bool pin_data) {
ArenaWrappedDBIter* iter = new ArenaWrappedDBIter();
Arena* arena = iter->GetArena();
auto mem = arena->AllocateAligned(sizeof(DBIter));
DBIter* db_iter =
new (mem) DBIter(env, ioptions, user_key_comparator, nullptr, sequence,
true, max_sequential_skip_in_iterations, version_number,
iterate_upper_bound, prefix_same_as_start);
iter->SetDBIter(db_iter);
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 21:08:30 +01:00
if (pin_data) {
iter->PinData();
}
return iter;
}
} // namespace rocksdb