rocksdb/table/cuckoo/cuckoo_table_reader.cc
Hui Xiao 06e593376c Group SST write in flush, compaction and db open with new stats (#11910)
Summary:
## Context/Summary
Similar to https://github.com/facebook/rocksdb/pull/11288, https://github.com/facebook/rocksdb/pull/11444, categorizing SST/blob file write according to different io activities allows more insight into the activity.

For that, this PR does the following:
- Tag different write IOs by passing down and converting WriteOptions to IOOptions
- Add new SST_WRITE_MICROS histogram in WritableFileWriter::Append() and breakdown FILE_WRITE_{FLUSH|COMPACTION|DB_OPEN}_MICROS

Some related code refactory to make implementation cleaner:
- Blob stats
   - Replace high-level write measurement with low-level WritableFileWriter::Append() measurement for BLOB_DB_BLOB_FILE_WRITE_MICROS. This is to make FILE_WRITE_{FLUSH|COMPACTION|DB_OPEN}_MICROS  include blob file. As a consequence, this introduces some behavioral changes on it, see HISTORY and db bench test plan below for more info.
   - Fix bugs where BLOB_DB_BLOB_FILE_SYNCED/BLOB_DB_BLOB_FILE_BYTES_WRITTEN include file failed to sync and bytes failed to write.
- Refactor WriteOptions constructor for easier construction with io_activity and rate_limiter_priority
- Refactor DBImpl::~DBImpl()/BlobDBImpl::Close() to bypass thread op verification
- Build table
   - TableBuilderOptions now includes Read/WriteOpitons so BuildTable() do not need to take these two variables
   - Replace the io_priority passed into BuildTable() with TableBuilderOptions::WriteOpitons::rate_limiter_priority. Similar for BlobFileBuilder.
This parameter is used for dynamically changing file io priority for flush, see  https://github.com/facebook/rocksdb/pull/9988?fbclid=IwAR1DtKel6c-bRJAdesGo0jsbztRtciByNlvokbxkV6h_L-AE9MACzqRTT5s for more
   - Update ThreadStatus::FLUSH_BYTES_WRITTEN to use io_activity to track flush IO in flush job and db open instead of io_priority

## Test
### db bench

Flush
```
./db_bench --statistics=1 --benchmarks=fillseq --num=100000 --write_buffer_size=100

rocksdb.sst.write.micros P50 : 1.830863 P95 : 4.094720 P99 : 6.578947 P100 : 26.000000 COUNT : 7875 SUM : 20377
rocksdb.file.write.flush.micros P50 : 1.830863 P95 : 4.094720 P99 : 6.578947 P100 : 26.000000 COUNT : 7875 SUM : 20377
rocksdb.file.write.compaction.micros P50 : 0.000000 P95 : 0.000000 P99 : 0.000000 P100 : 0.000000 COUNT : 0 SUM : 0
rocksdb.file.write.db.open.micros P50 : 0.000000 P95 : 0.000000 P99 : 0.000000 P100 : 0.000000 COUNT : 0 SUM : 0
```

compaction, db oopen
```
Setup: ./db_bench --statistics=1 --benchmarks=fillseq --num=10000 --disable_auto_compactions=1 -write_buffer_size=100 --db=../db_bench

Run:./db_bench --statistics=1 --benchmarks=compact  --db=../db_bench --use_existing_db=1

rocksdb.sst.write.micros P50 : 2.675325 P95 : 9.578788 P99 : 18.780000 P100 : 314.000000 COUNT : 638 SUM : 3279
rocksdb.file.write.flush.micros P50 : 0.000000 P95 : 0.000000 P99 : 0.000000 P100 : 0.000000 COUNT : 0 SUM : 0
rocksdb.file.write.compaction.micros P50 : 2.757353 P95 : 9.610687 P99 : 19.316667 P100 : 314.000000 COUNT : 615 SUM : 3213
rocksdb.file.write.db.open.micros P50 : 2.055556 P95 : 3.925000 P99 : 9.000000 P100 : 9.000000 COUNT : 23 SUM : 66
```

blob stats - just to make sure they aren't broken by this PR
```
Integrated Blob DB

Setup: ./db_bench --enable_blob_files=1 --statistics=1 --benchmarks=fillseq --num=10000 --disable_auto_compactions=1 -write_buffer_size=100 --db=../db_bench

Run:./db_bench --enable_blob_files=1 --statistics=1 --benchmarks=compact  --db=../db_bench --use_existing_db=1

pre-PR:
rocksdb.blobdb.blob.file.write.micros P50 : 7.298246 P95 : 9.771930 P99 : 9.991813 P100 : 16.000000 COUNT : 235 SUM : 1600
rocksdb.blobdb.blob.file.synced COUNT : 1
rocksdb.blobdb.blob.file.bytes.written COUNT : 34842

post-PR:
rocksdb.blobdb.blob.file.write.micros P50 : 2.000000 P95 : 2.829360 P99 : 2.993779 P100 : 9.000000 COUNT : 707 SUM : 1614
- COUNT is higher and values are smaller as it includes header and footer write
- COUNT is 3X higher due to each Append() count as one post-PR, while in pre-PR, 3 Append()s counts as one. See https://github.com/facebook/rocksdb/pull/11910/files#diff-32b811c0a1c000768cfb2532052b44dc0b3bf82253f3eab078e15ff201a0dabfL157-L164

rocksdb.blobdb.blob.file.synced COUNT : 1 (stay the same)
rocksdb.blobdb.blob.file.bytes.written COUNT : 34842 (stay the same)
```

```
Stacked Blob DB

Run: ./db_bench --use_blob_db=1 --statistics=1 --benchmarks=fillseq --num=10000 --disable_auto_compactions=1 -write_buffer_size=100 --db=../db_bench

pre-PR:
rocksdb.blobdb.blob.file.write.micros P50 : 12.808042 P95 : 19.674497 P99 : 28.539683 P100 : 51.000000 COUNT : 10000 SUM : 140876
rocksdb.blobdb.blob.file.synced COUNT : 8
rocksdb.blobdb.blob.file.bytes.written COUNT : 1043445

post-PR:
rocksdb.blobdb.blob.file.write.micros P50 : 1.657370 P95 : 2.952175 P99 : 3.877519 P100 : 24.000000 COUNT : 30001 SUM : 67924
- COUNT is higher and values are smaller as it includes header and footer write
- COUNT is 3X higher due to each Append() count as one post-PR, while in pre-PR, 3 Append()s counts as one. See https://github.com/facebook/rocksdb/pull/11910/files#diff-32b811c0a1c000768cfb2532052b44dc0b3bf82253f3eab078e15ff201a0dabfL157-L164

rocksdb.blobdb.blob.file.synced COUNT : 8 (stay the same)
rocksdb.blobdb.blob.file.bytes.written COUNT : 1043445 (stay the same)
```

###  Rehearsal CI stress test
Trigger 3 full runs of all our CI stress tests

###  Performance

Flush
```
TEST_TMPDIR=/dev/shm ./db_basic_bench_pre_pr --benchmark_filter=ManualFlush/key_num:524288/per_key_size:256 --benchmark_repetitions=1000
-- default: 1 thread is used to run benchmark; enable_statistics = true

Pre-pr: avg 507515519.3 ns
497686074,499444327,500862543,501389862,502994471,503744435,504142123,504224056,505724198,506610393,506837742,506955122,507695561,507929036,508307733,508312691,508999120,509963561,510142147,510698091,510743096,510769317,510957074,511053311,511371367,511409911,511432960,511642385,511691964,511730908,

Post-pr: avg 511971266.5 ns, regressed 0.88%
502744835,506502498,507735420,507929724,508313335,509548582,509994942,510107257,510715603,511046955,511352639,511458478,512117521,512317380,512766303,512972652,513059586,513804934,513808980,514059409,514187369,514389494,514447762,514616464,514622882,514641763,514666265,514716377,514990179,515502408,
```

Compaction
```
TEST_TMPDIR=/dev/shm ./db_basic_bench_{pre|post}_pr --benchmark_filter=ManualCompaction/comp_style:0/max_data:134217728/per_key_size:256/enable_statistics:1  --benchmark_repetitions=1000
-- default: 1 thread is used to run benchmark

Pre-pr: avg 495346098.30 ns
492118301,493203526,494201411,494336607,495269217,495404950,496402598,497012157,497358370,498153846

Post-pr: avg 504528077.20, regressed 1.85%. "ManualCompaction" include flush so the isolated regression for compaction should be around 1.85-0.88 = 0.97%
502465338,502485945,502541789,502909283,503438601,504143885,506113087,506629423,507160414,507393007
```

Put with WAL (in case passing WriteOptions slows down this path even without collecting SST write stats)
```
TEST_TMPDIR=/dev/shm ./db_basic_bench_pre_pr --benchmark_filter=DBPut/comp_style:0/max_data:107374182400/per_key_size:256/enable_statistics:1/wal:1  --benchmark_repetitions=1000
-- default: 1 thread is used to run benchmark

Pre-pr: avg 3848.10 ns
3814,3838,3839,3848,3854,3854,3854,3860,3860,3860

Post-pr: avg 3874.20 ns, regressed 0.68%
3863,3867,3871,3874,3875,3877,3877,3877,3880,3881
```

Pull Request resolved: https://github.com/facebook/rocksdb/pull/11910

Reviewed By: ajkr

Differential Revision: D49788060

Pulled By: hx235

fbshipit-source-id: 79e73699cda5be3b66461687e5147c2484fc5eff
2023-12-29 15:29:23 -08:00

414 lines
14 KiB
C++

// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
//
// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
#include "table/cuckoo/cuckoo_table_reader.h"
#include <algorithm>
#include <limits>
#include <string>
#include <utility>
#include <vector>
#include "memory/arena.h"
#include "options/cf_options.h"
#include "rocksdb/iterator.h"
#include "rocksdb/table.h"
#include "table/cuckoo/cuckoo_table_factory.h"
#include "table/get_context.h"
#include "table/internal_iterator.h"
#include "table/meta_blocks.h"
#include "util/coding.h"
namespace ROCKSDB_NAMESPACE {
namespace {
const uint64_t CACHE_LINE_MASK = ~((uint64_t)CACHE_LINE_SIZE - 1);
const uint32_t kInvalidIndex = std::numeric_limits<uint32_t>::max();
} // namespace
extern const uint64_t kCuckooTableMagicNumber;
CuckooTableReader::CuckooTableReader(
const ImmutableOptions& ioptions,
std::unique_ptr<RandomAccessFileReader>&& file, uint64_t file_size,
const Comparator* comparator,
uint64_t (*get_slice_hash)(const Slice&, uint32_t, uint64_t))
: file_(std::move(file)),
is_last_level_(false),
identity_as_first_hash_(false),
use_module_hash_(false),
num_hash_func_(0),
key_length_(0),
user_key_length_(0),
value_length_(0),
bucket_length_(0),
cuckoo_block_size_(0),
cuckoo_block_bytes_minus_one_(0),
table_size_(0),
ucomp_(comparator),
get_slice_hash_(get_slice_hash) {
if (!ioptions.allow_mmap_reads) {
status_ = Status::InvalidArgument("File is not mmaped");
return;
}
{
std::unique_ptr<TableProperties> props;
// TODO: plumb Env::IOActivity, Env::IOPriority
const ReadOptions read_options;
status_ =
ReadTableProperties(file_.get(), file_size, kCuckooTableMagicNumber,
ioptions, read_options, &props);
if (!status_.ok()) {
return;
}
table_props_ = std::move(props);
}
auto& user_props = table_props_->user_collected_properties;
auto hash_funs = user_props.find(CuckooTablePropertyNames::kNumHashFunc);
if (hash_funs == user_props.end()) {
status_ = Status::Corruption("Number of hash functions not found");
return;
}
num_hash_func_ = *reinterpret_cast<const uint32_t*>(hash_funs->second.data());
auto unused_key = user_props.find(CuckooTablePropertyNames::kEmptyKey);
if (unused_key == user_props.end()) {
status_ = Status::Corruption("Empty bucket value not found");
return;
}
unused_key_ = unused_key->second;
key_length_ = static_cast<uint32_t>(table_props_->fixed_key_len);
auto user_key_len = user_props.find(CuckooTablePropertyNames::kUserKeyLength);
if (user_key_len == user_props.end()) {
status_ = Status::Corruption("User key length not found");
return;
}
user_key_length_ =
*reinterpret_cast<const uint32_t*>(user_key_len->second.data());
auto value_length = user_props.find(CuckooTablePropertyNames::kValueLength);
if (value_length == user_props.end()) {
status_ = Status::Corruption("Value length not found");
return;
}
value_length_ =
*reinterpret_cast<const uint32_t*>(value_length->second.data());
bucket_length_ = key_length_ + value_length_;
auto hash_table_size =
user_props.find(CuckooTablePropertyNames::kHashTableSize);
if (hash_table_size == user_props.end()) {
status_ = Status::Corruption("Hash table size not found");
return;
}
table_size_ =
*reinterpret_cast<const uint64_t*>(hash_table_size->second.data());
auto is_last_level = user_props.find(CuckooTablePropertyNames::kIsLastLevel);
if (is_last_level == user_props.end()) {
status_ = Status::Corruption("Is last level not found");
return;
}
is_last_level_ = *reinterpret_cast<const bool*>(is_last_level->second.data());
auto identity_as_first_hash =
user_props.find(CuckooTablePropertyNames::kIdentityAsFirstHash);
if (identity_as_first_hash == user_props.end()) {
status_ = Status::Corruption("identity as first hash not found");
return;
}
identity_as_first_hash_ =
*reinterpret_cast<const bool*>(identity_as_first_hash->second.data());
auto use_module_hash =
user_props.find(CuckooTablePropertyNames::kUseModuleHash);
if (use_module_hash == user_props.end()) {
status_ = Status::Corruption("hash type is not found");
return;
}
use_module_hash_ =
*reinterpret_cast<const bool*>(use_module_hash->second.data());
auto cuckoo_block_size =
user_props.find(CuckooTablePropertyNames::kCuckooBlockSize);
if (cuckoo_block_size == user_props.end()) {
status_ = Status::Corruption("Cuckoo block size not found");
return;
}
cuckoo_block_size_ =
*reinterpret_cast<const uint32_t*>(cuckoo_block_size->second.data());
cuckoo_block_bytes_minus_one_ = cuckoo_block_size_ * bucket_length_ - 1;
// TODO: rate limit reads of whole cuckoo tables.
status_ = file_->Read(IOOptions(), 0, static_cast<size_t>(file_size),
&file_data_, nullptr, nullptr);
}
Status CuckooTableReader::Get(const ReadOptions& /*readOptions*/,
const Slice& key, GetContext* get_context,
const SliceTransform* /* prefix_extractor */,
bool /*skip_filters*/) {
assert(key.size() == key_length_ + (is_last_level_ ? 8 : 0));
Slice user_key = ExtractUserKey(key);
for (uint32_t hash_cnt = 0; hash_cnt < num_hash_func_; ++hash_cnt) {
uint64_t offset =
bucket_length_ * CuckooHash(user_key, hash_cnt, use_module_hash_,
table_size_, identity_as_first_hash_,
get_slice_hash_);
const char* bucket = &file_data_.data()[offset];
for (uint32_t block_idx = 0; block_idx < cuckoo_block_size_;
++block_idx, bucket += bucket_length_) {
if (ucomp_->Equal(Slice(unused_key_.data(), user_key.size()),
Slice(bucket, user_key.size()))) {
return Status::OK();
}
// Here, we compare only the user key part as we support only one entry
// per user key and we don't support snapshot.
if (ucomp_->Equal(user_key, Slice(bucket, user_key.size()))) {
Slice value(bucket + key_length_, value_length_);
if (is_last_level_) {
// Sequence number is not stored at the last level, so we will use
// kMaxSequenceNumber since it is unknown. This could cause some
// transactions to fail to lock a key due to known sequence number.
// However, it is expected for anyone to use a CuckooTable in a
// TransactionDB.
get_context->SaveValue(value, kMaxSequenceNumber);
} else {
Slice full_key(bucket, key_length_);
ParsedInternalKey found_ikey;
Status s = ParseInternalKey(full_key, &found_ikey,
false /* log_err_key */); // TODO
if (!s.ok()) {
return s;
}
bool dont_care __attribute__((__unused__));
get_context->SaveValue(found_ikey, value, &dont_care);
}
// We don't support merge operations. So, we return here.
return Status::OK();
}
}
}
return Status::OK();
}
void CuckooTableReader::Prepare(const Slice& key) {
// Prefetch the first Cuckoo Block.
Slice user_key = ExtractUserKey(key);
uint64_t addr =
reinterpret_cast<uint64_t>(file_data_.data()) +
bucket_length_ * CuckooHash(user_key, 0, use_module_hash_, table_size_,
identity_as_first_hash_, nullptr);
uint64_t end_addr = addr + cuckoo_block_bytes_minus_one_;
for (addr &= CACHE_LINE_MASK; addr < end_addr; addr += CACHE_LINE_SIZE) {
PREFETCH(reinterpret_cast<const char*>(addr), 0, 3);
}
}
class CuckooTableIterator : public InternalIterator {
public:
explicit CuckooTableIterator(CuckooTableReader* reader);
// No copying allowed
CuckooTableIterator(const CuckooTableIterator&) = delete;
void operator=(const Iterator&) = delete;
~CuckooTableIterator() override = default;
bool Valid() const override;
void SeekToFirst() override;
void SeekToLast() override;
void Seek(const Slice& target) override;
void SeekForPrev(const Slice& target) override;
void Next() override;
void Prev() override;
Slice key() const override;
Slice value() const override;
Status status() const override { return Status::OK(); }
void InitIfNeeded();
private:
struct BucketComparator {
BucketComparator(const Slice& file_data, const Comparator* ucomp,
uint32_t bucket_len, uint32_t user_key_len,
const Slice& target = Slice())
: file_data_(file_data),
ucomp_(ucomp),
bucket_len_(bucket_len),
user_key_len_(user_key_len),
target_(target) {}
bool operator()(const uint32_t first, const uint32_t second) const {
const char* first_bucket = (first == kInvalidIndex)
? target_.data()
: &file_data_.data()[first * bucket_len_];
const char* second_bucket =
(second == kInvalidIndex) ? target_.data()
: &file_data_.data()[second * bucket_len_];
return ucomp_->Compare(Slice(first_bucket, user_key_len_),
Slice(second_bucket, user_key_len_)) < 0;
}
private:
const Slice file_data_;
const Comparator* ucomp_;
const uint32_t bucket_len_;
const uint32_t user_key_len_;
const Slice target_;
};
const BucketComparator bucket_comparator_;
void PrepareKVAtCurrIdx();
CuckooTableReader* reader_;
bool initialized_;
// Contains a map of keys to bucket_id sorted in key order.
std::vector<uint32_t> sorted_bucket_ids_;
// We assume that the number of items can be stored in uint32 (4 Billion).
uint32_t curr_key_idx_;
Slice curr_value_;
IterKey curr_key_;
};
CuckooTableIterator::CuckooTableIterator(CuckooTableReader* reader)
: bucket_comparator_(reader->file_data_, reader->ucomp_,
reader->bucket_length_, reader->user_key_length_),
reader_(reader),
initialized_(false),
curr_key_idx_(kInvalidIndex) {
sorted_bucket_ids_.clear();
curr_value_.clear();
curr_key_.Clear();
}
void CuckooTableIterator::InitIfNeeded() {
if (initialized_) {
return;
}
sorted_bucket_ids_.reserve(
static_cast<size_t>(reader_->GetTableProperties()->num_entries));
uint64_t num_buckets = reader_->table_size_ + reader_->cuckoo_block_size_ - 1;
assert(num_buckets < kInvalidIndex);
const char* bucket = reader_->file_data_.data();
for (uint32_t bucket_id = 0; bucket_id < num_buckets; ++bucket_id) {
if (Slice(bucket, reader_->key_length_) != Slice(reader_->unused_key_)) {
sorted_bucket_ids_.push_back(bucket_id);
}
bucket += reader_->bucket_length_;
}
assert(sorted_bucket_ids_.size() ==
reader_->GetTableProperties()->num_entries);
std::sort(sorted_bucket_ids_.begin(), sorted_bucket_ids_.end(),
bucket_comparator_);
curr_key_idx_ = kInvalidIndex;
initialized_ = true;
}
void CuckooTableIterator::SeekToFirst() {
InitIfNeeded();
curr_key_idx_ = 0;
PrepareKVAtCurrIdx();
}
void CuckooTableIterator::SeekToLast() {
InitIfNeeded();
curr_key_idx_ = static_cast<uint32_t>(sorted_bucket_ids_.size()) - 1;
PrepareKVAtCurrIdx();
}
void CuckooTableIterator::Seek(const Slice& target) {
InitIfNeeded();
const BucketComparator seek_comparator(
reader_->file_data_, reader_->ucomp_, reader_->bucket_length_,
reader_->user_key_length_, ExtractUserKey(target));
auto seek_it =
std::lower_bound(sorted_bucket_ids_.begin(), sorted_bucket_ids_.end(),
kInvalidIndex, seek_comparator);
curr_key_idx_ =
static_cast<uint32_t>(std::distance(sorted_bucket_ids_.begin(), seek_it));
PrepareKVAtCurrIdx();
}
void CuckooTableIterator::SeekForPrev(const Slice& /*target*/) {
// Not supported
assert(false);
}
bool CuckooTableIterator::Valid() const {
return curr_key_idx_ < sorted_bucket_ids_.size();
}
void CuckooTableIterator::PrepareKVAtCurrIdx() {
if (!Valid()) {
curr_value_.clear();
curr_key_.Clear();
return;
}
uint32_t id = sorted_bucket_ids_[curr_key_idx_];
const char* offset =
reader_->file_data_.data() + id * reader_->bucket_length_;
if (reader_->is_last_level_) {
// Always return internal key.
curr_key_.SetInternalKey(Slice(offset, reader_->user_key_length_), 0,
kTypeValue);
} else {
curr_key_.SetInternalKey(Slice(offset, reader_->key_length_));
}
curr_value_ = Slice(offset + reader_->key_length_, reader_->value_length_);
}
void CuckooTableIterator::Next() {
if (!Valid()) {
curr_value_.clear();
curr_key_.Clear();
return;
}
++curr_key_idx_;
PrepareKVAtCurrIdx();
}
void CuckooTableIterator::Prev() {
if (curr_key_idx_ == 0) {
curr_key_idx_ = static_cast<uint32_t>(sorted_bucket_ids_.size());
}
if (!Valid()) {
curr_value_.clear();
curr_key_.Clear();
return;
}
--curr_key_idx_;
PrepareKVAtCurrIdx();
}
Slice CuckooTableIterator::key() const {
assert(Valid());
return curr_key_.GetInternalKey();
}
Slice CuckooTableIterator::value() const {
assert(Valid());
return curr_value_;
}
InternalIterator* CuckooTableReader::NewIterator(
const ReadOptions& /*read_options*/,
const SliceTransform* /* prefix_extractor */, Arena* arena,
bool /*skip_filters*/, TableReaderCaller /*caller*/,
size_t /*compaction_readahead_size*/, bool /* allow_unprepared_value */) {
if (!status().ok()) {
return NewErrorInternalIterator<Slice>(
Status::Corruption("CuckooTableReader status is not okay."), arena);
}
CuckooTableIterator* iter;
if (arena == nullptr) {
iter = new CuckooTableIterator(this);
} else {
auto iter_mem = arena->AllocateAligned(sizeof(CuckooTableIterator));
iter = new (iter_mem) CuckooTableIterator(this);
}
return iter;
}
size_t CuckooTableReader::ApproximateMemoryUsage() const { return 0; }
} // namespace ROCKSDB_NAMESPACE