Summary:
When timestamp is enabled, key comparison should take this into account.
In `BlockBasedTableReader::Get()`, `BlockBasedTableReader::MultiGet()`,
assume the target key is `key`, and the timestamp upper bound is `ts`.
The highest key in current block is (key, ts1), while the lowest key in next
block is (key, ts2).
If
```
ts1 > ts > ts2
```
then
```
(key, ts1) < (key, ts) < (key, ts2)
```
It can be shown that if `Compare()` is used, then we will mistakenly skip the next
block. Instead, we should use `CompareWithoutTimestamp()`.

The majority of this PR makes some existing tests in `db_with_timestamp_basic_test.cc`
parameterized so that different index types can be tested. A new unit test is
also added for more coverage.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8062

Test Plan: make check

Reviewed By: ltamasi

Differential Revision: D27057557

Pulled By: riversand963

fbshipit-source-id: c1062fa7c159ed600a1ad7e461531d52265021f1

Summary:
As title.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8054

Test Plan: make check

Reviewed By: mrambacher

Differential Revision: D27017955

Pulled By: riversand963

fbshipit-source-id: 829497d507bc89afbe982f8a8cf3555e52fd7098

Summary:
For performance purposes, the lower level routines were changed to use a SystemClock* instead of a std::shared_ptr<SystemClock>.  The shared ptr has some performance degradation on certain hardware classes.

For most of the system, there is no risk of the pointer being deleted/invalid because the shared_ptr will be stored elsewhere.  For example, the ImmutableDBOptions stores the Env which has a std::shared_ptr<SystemClock> in it.  The SystemClock* within the ImmutableDBOptions is essentially a "short cut" to gain access to this constant resource.

There were a few classes (PeriodicWorkScheduler?) where the "short cut" property did not hold.  In those cases, the shared pointer was preserved.

Using db_bench readrandom perf_level=3 on my EC2 box, this change performed as well or better than 6.17:

6.17: readrandom   :      28.046 micros/op 854902 ops/sec;   61.3 MB/s (355999 of 355999 found)
6.18: readrandom   :      32.615 micros/op 735306 ops/sec;   52.7 MB/s (290999 of 290999 found)
PR: readrandom   :      27.500 micros/op 871909 ops/sec;   62.5 MB/s (367999 of 367999 found)

(Note that the times for 6.18 are prior to revert of the SystemClock).

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8033

Reviewed By: pdillinger

Differential Revision: D27014563

Pulled By: mrambacher

fbshipit-source-id: ad0459eba03182e454391b5926bf5cdd45657b67

Summary:
CompactionDeletionTriggerReopen was observed to be flaky recently:
https://app.circleci.com/pipelines/github/facebook/rocksdb/6030/workflows/787af4f3-b9f7-4645-8e8d-1fb0ebf05539/jobs/101451.

I went through it and the related tests and arrived at different
conclusions on what constraints we can expect on DB size. Some
constraints got looser and some got tighter. The particular constraint
that flaked got a lot looser so at least the flake linked above would have been prevented.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8036

Reviewed By: riversand963

Differential Revision: D26862566

Pulled By: ajkr

fbshipit-source-id: 3512b86b4fb41aeecae32e1c7382c03916d88d88

Summary:
`DBTest.GetLiveBlobFiles` and `ObsoleteFilesTest.BlobFiles` both modify the
current `Version` in their setup phase, implicitly assuming that no other
threads would touch the `Version` while this is happening. The periodic
stats dumper thread violates this assumption; the patch fixes this by
disabling it in the affected test cases. (Note: the data race is
harmless in the sense that it only affects test code.)

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8055

Test Plan:
```
COMPILE_WITH_TSAN=1 make db_test -j24
gtest-parallel --repeat=10000 ./db_test --gtest_filter="*GetLiveBlobFiles"
COMPILE_WITH_TSAN=1 make obsolete_files_test -j24
gtest-parallel --repeat=10000 ./obsolete_files_test --gtest_filter="*BlobFiles"
```

Reviewed By: riversand963

Differential Revision: D27022715

Pulled By: ltamasi

fbshipit-source-id: b6cc77ed63d8bc1cbe0603522ff1a572182fc9ab

Summary:
This is for cases that do not meet the Facebook criteria for
SKIP (see new comments). Also made ROCKSDB_GTEST_{SKIP,BYPASS} print the
message because gtest doesn't ever seem to.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8048

Test Plan: manual inspection of ./ribbon_test output, CI

Reviewed By: mrambacher

Differential Revision: D26953688

Pulled By: pdillinger

fbshipit-source-id: c914eaffe7d419db6ab90a193d474531e23582e5

Summary:
a trial gtest upgrade discovered some parameterized tests missing instantiation. By some miracle, they still pass.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8051

Test Plan: thisisthetest

Reviewed By: mrambacher

Differential Revision: D27003684

Pulled By: pdillinger

fbshipit-source-id: cde1cab1551fb282f67d462d46574bd30bd5e61f

Summary:
This API can be used for things like determining how much space
can be freed up by deleting a particular backup, etc.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8042

Test Plan:
validation of the API added to many existing backup unit
tests

Reviewed By: mrambacher

Differential Revision: D26936577

Pulled By: pdillinger

fbshipit-source-id: f0bbd90f0917b9781a6837652fb4616d9247816a

Summary:
This PR does the following:

- Enable backward iteration for keys with user-defined timestamp. Note that merge, single delete, range delete are not supported yet.
- Introduces a new helper API `Comparator::EqualWithoutTimestamp()`.
- Fix a typo in `SetTimestamp()`.
- Add/update unit tests

Run db_bench (built with DEBUG_LEVEL=0) to demonstrate that no overhead is introduced for CPU-intensive workloads with a lot of `Prev()`. Also provided results of iterating keys with timestamps.

1. Disable timestamp, run:
```
./db_bench -db=/dev/shm/rocksdb -disable_wal=1 -benchmarks=fillseq,seekrandom[-W1-X6] -reverse_iterator=1 -seek_nexts=5
```
Results:
> Baseline
> - seekrandom [AVG    6 runs] : 96115 ops/sec;   53.2 MB/sec
> - seekrandom [MEDIAN 6 runs] : 98075 ops/sec;   54.2 MB/sec
>
> This PR
> - seekrandom [AVG    6 runs] : 95521 ops/sec;   52.8 MB/sec
> - seekrandom [MEDIAN 6 runs] : 96338 ops/sec;   53.3 MB/sec

2. Enable timestamp, run:
```
./db_bench -user_timestamp_size=8  -db=/dev/shm/rocksdb -disable_wal=1 -benchmarks=fillseq,seekrandom[-W1-X6] -reverse_iterator=1 -seek_nexts=5
```
Result:
> Baseline: not supported
>
> This PR
> - seekrandom [AVG    6 runs] : 90514 ops/sec;   50.1 MB/sec
> - seekrandom [MEDIAN 6 runs] : 90834 ops/sec;   50.2 MB/sec

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8035

Reviewed By: ltamasi

Differential Revision: D26926668

Pulled By: riversand963

fbshipit-source-id: 95330cc2242397c03e09d29e5417dfb0adc98ef5

Summary:
This PR

- adds a class `ManifestTailer` that inherits from `VersionEditHandlerPointInTime`. `ManifestTailer::Iterate()` can be called multiple times to tail the primary instance's MANIFEST and apply the changes to the secondary,
- updates the implementation of `ReactiveVersionSet::ReadAndApply` to use this class,
- removes unused code in version_set.cc,
- updates existing tests, e.g. removing deleted sync points from unit tests,
- adds a new test to address the bug in https://github.com/facebook/rocksdb/issues/7815.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7998

Test Plan:
make check
Existing and newly-added tests in version_set_test.cc and db_secondary_test.cc

Reviewed By: jay-zhuang

Differential Revision: D26926641

Pulled By: riversand963

fbshipit-source-id: 8d4dd15db0ba863c213f743e33b5a207e948c980

Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/7893

Reviewed By: ajkr

Differential Revision: D26050966

Pulled By: jay-zhuang

fbshipit-source-id: 9df2bb65d82defd7fad49d5369979b03e22d39c2

Summary:
The variable `byteCompressionType` is only assigned values of primitive type and is never 'null', but it is declared with the boxed type 'Byte'.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7981

Reviewed By: ajkr

Differential Revision: D26546600

Pulled By: jay-zhuang

fbshipit-source-id: 07b579cdfcfc2262a448ca3626e216416fd05892

Summary:
OS: Ubuntu 14.04
Compiler: GCC 4.9.4
Compile error:
```
db/forward_iterator.cc:996:62: error: declaration of ‘key’ shadows a member of 'this' [-Werror=shadow]
   auto cmp = [&](const FileMetaData* f, const Slice& key) -> bool {
```

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7908

Reviewed By: jay-zhuang

Differential Revision: D26899986

Pulled By: ajkr

fbshipit-source-id: 66b0b97aefd0f13a085e063491f8207366a9f848

Summary:
This PR adds support for custom file systems to ldb and sst_dump by adding command line options for specifying --fs_uri and --backup_fs uri (for ldb backup/restore commands). fs_uri is already supported in db_bench and db_stress, and there is already support in ldb and db stress for specifying customized envs.

The PR also fixes what looks like a bug in the ldb backup/restore commands. As it is right now, backups can only be made from and to the same environment/file system which does not seem to be the intended behavior. This PR makes it possible to do/restore backups between different envs/file systems.

Example:
`./ldb backup --fs_uri=zenfs://dev:nvme2n1 --backup_fs_uri=posix:// --backup_dir=/tmp/my_rocksdb_backup  --db=rocksdbtest/dbbench
`

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8010

Reviewed By: jay-zhuang

Differential Revision: D26904654

Pulled By: ajkr

fbshipit-source-id: 9b695ed8b944fcc6b27c4daaa9f52e87ee2c1fb4

Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/7994

Reviewed By: jay-zhuang

Differential Revision: D26904603

Pulled By: ajkr

fbshipit-source-id: 0af92a51de895b40c7faaa4f0870b3f63279fe21

Summary:
Removed confusing, awkward, and undocumented internal API
ReadOneLine and replaced with very simple LineFileReader.

In refactoring backupable_db.cc, this has the side benefit of
removing the arbitrary cap on the size of backup metadata files.

Also added Status::MustCheck to make it easy to mark a Status as
"must check." Using this, I can ensure that after
LineFileReader::ReadLine returns false the caller checks GetStatus().

Also removed some excessive conditional compilation in status.h

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8026

Test Plan: added unit test, and running tests with ASSERT_STATUS_CHECKED

Reviewed By: mrambacher

Differential Revision: D26831687

Pulled By: pdillinger

fbshipit-source-id: ef749c265a7a26bb13cd44f6f0f97db2955f6f0f

Summary:
New comment for share_files_with_checksum:
// Only used if share_table_files is set to true. Setting to false is
// DEPRECATED and potentially dangerous because in that case BackupEngine
// can lose data if backing up databases with distinct or divergent
// history, for example if restoring from a backup other than the latest,
// writing to the DB, and creating another backup. Setting to true (default)
// prevents these issues by ensuring that different table files (SSTs) with
// the same number are treated as distinct. See
// share_files_with_checksum_naming and ShareFilesNaming.

I have also removed interim option kFlagMatchInterimNaming, which is no
longer needed and was never needed for correct+compatible operation
(just performance).

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8020

Test Plan:
tests updated. Backward+forward compatibility verified with
SHORT_TEST=1 check_format_compatible.sh. ldb uses default backup
options, and I manually verified shared_checksum in
/tmp/rocksdb_format_compatible_peterd/bak/current/ after run.

Reviewed By: ajkr

Differential Revision: D26786331

Pulled By: pdillinger

fbshipit-source-id: 36f968dfef1f5cacbd65154abe1d846151a55130

Summary:
Haven't seen any production issues with new Bloom filter and
it's now > 1 year old (added in 6.6.0).

Updated check_format_compatible.sh and HISTORY.md

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8017

Test Plan: tests updated (or prior bugs fixed)

Reviewed By: ajkr

Differential Revision: D26762197

Pulled By: pdillinger

fbshipit-source-id: 0e755c46b443087c1544da0fd545beb9c403d1c2

Summary:
`TtlDB.open()`: missing space after 'column'
`AdvancedColumnFamilyOptionsInterface.setLevelCompactionDynamicLevelBytes()`: missing space after 'cause'

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7982

Reviewed By: ajkr

Differential Revision: D26546632

Pulled By: jay-zhuang

fbshipit-source-id: 885dedcaa2200842764fbac9ce3766d54e1c8914

Summary:
Add ${ARTIFACT_SUFFIX} to benchmark tool names to enable differentiating jemalloc and non-jemalloc versions.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8016

Reviewed By: jay-zhuang

Differential Revision: D26907007

Pulled By: ajkr

fbshipit-source-id: 78d3b3372b5454d52d5b663ea982135ea9cf7bf8

Summary:
## 1. Bug description:

When RocksDB Checkpoint, it may be stuck in `WaitUntilFlushWouldNotStallWrites` method.

## 2. Simple analysis of the reasons:

### 2.1 Configuration parameters:

```yaml
Compaction Style : Universal

max_write_buffer_number : 4
min_write_buffer_number_to_merge : 3
```

Checkpoint is usually very fast. When the Checkpoint is executed, `WaitUntilFlushWouldNotStallWrites` is called. If there are 2 Immutable MemTables, which are less than `min_write_buffer_number_to_merge`, they will not be flushed. But will enter this code.

```c++
// method: GetWriteStallConditionAndCause
if (mutable_cf_options.max_write_buffer_number> 3 &&
              num_unflushed_memtables >=
                  mutable_cf_options.max_write_buffer_number-1) {
     return {WriteStallCondition::kDelayed, WriteStallCause::kMemtableLimit};
}
```

code link: https://github.com/facebook/rocksdb/blob/fbed72f03c3d9e4fdca3e5993587ef2559ba6ab9/db/column_family.cc#L847

Checkpoint thought there was a FlushJob, but it didn't. So will always wait.

### 2.2 solution:

Increase the restriction: the `number of Immutable MemTable` >= `min_write_buffer_number_to_merge will wait`.

If there are other better solutions, you can correct me.

### 2.3 Code that can reproduce the problem:

https://github.com/1996fanrui/fanrui-learning/blob/flink-1.12/module-java/src/main/java/com/dream/rocksdb/RocksDBCheckpointStuck.java

## 3. Interesting point

This bug will be triggered only when `the number of sorted runs >= level0_file_num_compaction_trigger`.

Because there is a break in WaitUntilFlushWouldNotStallWrites.

```c++
if (cfd->imm()->NumNotFlushed() <
        cfd->ioptions()->min_write_buffer_number_to_merge &&
    vstorage->l0_delay_trigger_count() <
        mutable_cf_options.level0_file_num_compaction_trigger) {
  break;
}
```

code link: https://github.com/facebook/rocksdb/blob/fbed72f03c3d9e4fdca3e5993587ef2559ba6ab9/db/db_impl/db_impl_compaction_flush.cc#L1974

Universal may have `l0_delay_trigger_count() >= level0_file_num_compaction_trigger`, so this bug is triggered.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7921

Reviewed By: jay-zhuang

Differential Revision: D26900559

Pulled By: ajkr

fbshipit-source-id: 133c1252dad7393753f04a47590b68c7d8e670df

Summary:
concurrnet -> concurrent

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8001

Reviewed By: ajkr

Differential Revision: D26659381

Pulled By: riversand963

fbshipit-source-id: 890d102d1cf836ed3b183da66d3d56a3158017d0

Summary:
On x86_64, this makes the struct 8 bytes smaller, so creating a PerfStepTimer on the stack will use slightly less stack space.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7931

Reviewed By: jay-zhuang

Differential Revision: D26529470

Pulled By: ajkr

fbshipit-source-id: bbe2e843167152ffa05a5946f1add6621c9849f7

Summary:
We were seeing intermittent `ld` failures due to `No space left on device` such as https://travis-ci.org/github/facebook/rocksdb/jobs/761905070.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8043

Reviewed By: pdillinger

Differential Revision: D26889711

Pulled By: ajkr

fbshipit-source-id: 010b7617d339bddc30026586bfde41539632fb2d

Summary:
For some reason I still cannot figure out, the manual flush in this test
was sometimes producing a third tiny file. I saw it a bunch of times on
ppc64le, but even running a qemu system with that architecture (and
playing with various other options) could not repro. However we did get
an instrumented Travis run to confirm the problem is indeed a third tiny
file - https://travis-ci.org/github/facebook/rocksdb/jobs/761986592. We
can avoid it by filling memtables less full and using manual flush.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8044

Reviewed By: akankshamahajan15

Differential Revision: D26892635

Pulled By: ajkr

fbshipit-source-id: 775c04176931cf01d07cc78fb82cfe3a11beebcf

Summary:
I recently discovered the confusing, undocumented semantics of
Read() functions in the FileSystem and Env APIs. I have added
clarification to the best of my reverse-engineered understanding, and
made a note in HISTORY.md for implementors to check their
implementations, as a subtly non-adherent implementation could lead to
RocksDB quietly ignoring some portion of a file.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8029

Test Plan: no code changes

Reviewed By: anand1976

Differential Revision: D26831698

Pulled By: pdillinger

fbshipit-source-id: 208f97ff6037bc13bb2ef360b987c2640c79bd03

Summary:
The patch does the following:
1) Exposes the amount of data (number of bytes) read from blob files from
`BlobFileReader::GetBlob` / `Version::GetBlob`.
2) Tracks the total number and size of blobs read from blob files during a
compaction (due to garbage collection or compaction filter usage) in
`CompactionIterationStats` and propagates this data to
`InternalStats::CompactionStats` / `CompactionJobStats`.
3) Updates the formulae for write amplification calculations to include the
amount of data read from blob files.
4) Extends the compaction stats dump with a new column `Rblob(GB)` and
a new line containing the total number and size of blob files in the current
`Version` to complement the information about the shape and size of the LSM tree
that's already there.
5) Updates `CompactionJobStats` so that the number of files and amount of data
written by a compaction are broken down per file type (i.e. table/blob file).

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8022

Test Plan: Ran `make check` and `db_bench`.

Reviewed By: riversand963

Differential Revision: D26801199

Pulled By: ltamasi

fbshipit-source-id: 28a5f072048a702643b28cb5971b4099acabbfb2

Summary:
This PR adds SetBufferSize() to the WriteBufferManager object.  This enables user code to adjust the global budget for write_buffers based upon other memory conditions such as growth in table reader memory as the dataset grows.

The buffer_size_ member variable is now atomic to match design of other changeable size_t members within WriteBufferManager.

This change is useful as is.  However, this change is also essential if someone decides they wanted to enable db_write_buffer_size modifications through the DB::SetOptions() API, i.e. no waste taking this as is.

Any format / spacing changes are due to clang-format as required by check-in automation.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7961

Reviewed By: ajkr

Differential Revision: D26639075

Pulled By: akankshamahajan15

fbshipit-source-id: 0604348caf092d35f44e85715331dc920e5c1033

Summary:
When changing db iterator direction, we may perform a reseek.
Therefore, we should bump the NUMBER_OF_RESEEKS_IN_ITERATION counter.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8015

Test Plan: make check

Reviewed By: ltamasi

Differential Revision: D26755415

Pulled By: riversand963

fbshipit-source-id: 211f51f1a454bcda768fc46c0dce51edeb7f05fe

Summary:
* Adds backup/restore forward/backward compatibility testing
* Adds forward/backward compatibility testing to sst ingestion
* More structure sharing and comments for the lists of branches
comprising each group
* Less reliant on invariants between groups with de-duplication logic
* Restructured for n+1 branch checkout+build steps rather than something
like 3n. Should be much faster despite more checks.

And to make manual runs easier

* On success, restores working trees to original working branch (aborts
early if uncommitted changes) and deletes temporary branch & remote
* Adds SHORT_TEST=1 mode that uses only the oldest version for each
* Adds USE_SSH=1 to use ssh instead of https for github
group

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8012

Test Plan:
a number of manual tests, mostly with SHORT_TEST=1. Using one
version older for any of the groups (except I didn't check
db_backward_only_refs) fails. Changing default format_version to 5
(planned) without updating this script fails as it should, and passes
with appropriate update. Full local run passed (had to remove "2.7.fb.branch"
due to compiler issues, also before this change).

Reviewed By: riversand963

Differential Revision: D26735840

Pulled By: pdillinger

fbshipit-source-id: 1320c22de5674760657e385aa42df9fade8b6fff

Summary:
The patch breaks down the "bytes written" (as well as the "number of output files")
compaction statistics into two, so the values are logged separately for table files
and blob files in the info log, and are shown in separate columns (`Write(GB)` for table
files, `Wblob(GB)` for blob files) when the compaction statistics are dumped.
This will also come in handy for fixing the write amplification statistics, which currently
do not consider the amount of data read from blob files during compaction. (This will
be fixed by an upcoming patch.)

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8013

Test Plan: Ran `make check` and `db_bench`.

Reviewed By: riversand963

Differential Revision: D26742156

Pulled By: ltamasi

fbshipit-source-id: 31d18ee8f90438b438ca7ed1ea8cbd92114442d5

Summary:
The checkpointing logic supports passing file level checksums
to the copy_file_cb callback function which is used by the backup code
for detecting corruption during file copies.
However, this is currently implemented only for table files.

This PR extends the checksum retrieval to blob files as well.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8003

Test Plan: Add new test units

Reviewed By: ltamasi

Differential Revision: D26680701

Pulled By: akankshamahajan15

fbshipit-source-id: 1bd1e2464df6e9aa31091d35b8c72786d94cd1c5

Summary:
As title.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8011

Test Plan:
```
./db_bench -enable_blob_files=1 -use_keep_filter=1 -disable_auto_compactions=1
/db_stress -enable_blob_files=1 -enable_compaction_filter=1 -acquire_snapshot_one_in=0 -compact_range_one_in=0 -iterpercent=0 -test_batches_snapshots=0 -readpercent=10 -prefixpercent=20 -writepercent=55 -delpercent=15 -continuous_verification_interval=0
```

Reviewed By: ltamasi

Differential Revision: D26736061

Pulled By: riversand963

fbshipit-source-id: 1c7834903c28431ce23324c4f259ed71255614e2

Summary:
This is likely a temp fix before we figure out a better way.

PerfStepTimer is used intensively in certain benchmarking/testings. https://github.com/facebook/rocksdb/issues/7858 stores a `shared_ptr` to system clock in PerfStepTimer which gets created each time a `PerfStepTimer` object is created. The atomic operations in `shared_ptr` may add overhead in CPU cycles. Therefore, we change it back to a raw `SystemClock*` for now.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8006

Test Plan: make check

Reviewed By: pdillinger

Differential Revision: D26703560

Pulled By: riversand963

fbshipit-source-id: 519d0769b28da2334bea7d86c848fcc26ee8a17f

Summary:
This change only affects non-schema-critical aspects of the production candidate Ribbon filter. Specifically, it refines choice of internal configuration parameters based on inputs. The changes are minor enough that the schema tests in bloom_test, some of which depend on this, are unaffected. There are also some minor optimizations and refactorings.

This would be a schema change for "smash" Ribbon, to fix some known issues with small filters, but "smash" Ribbon is not accessible in public APIs. Unit test CompactnessAndBacktrackAndFpRate updated to test small and medium-large filters. Run with --thoroughness=100 or so for much better detection power (not appropriate for continuous regression testing).

Homogenous Ribbon:
This change adds internally a Ribbon filter variant we call Homogeneous Ribbon, in collaboration with Stefan Walzer. The expected "result" value for every key is zero, instead of computed from a hash. Entropy for queries not to be false positives comes from free variables ("overhead") in the solution structure, which are populated pseudorandomly. Construction is slightly faster for not tracking result values, and never fails. Instead, FP rate can jump up whenever and whereever entries are packed too tightly. For small structures, we can choose overhead to make this FP rate jump unlikely, as seen in updated unit test CompactnessAndBacktrackAndFpRate.

Unlike standard Ribbon, Homogeneous Ribbon seems to scale to arbitrary number of keys when accepting an FP rate penalty for small pockets of high FP rate in the structure. For example, 64-bit ribbon with 8 solution columns and 10% allocated space overhead for slots seems to achieve about 10.5% space overhead vs. information-theoretic minimum based on its observed FP rate with expected pockets of degradation. (FP rate is close to 1/256.) If targeting a higher FP rate with fewer solution columns, Homogeneous Ribbon can be even more space efficient, because the penalty from degradation is relatively smaller. If targeting a lower FP rate, Homogeneous Ribbon is less space efficient, as more allocated overhead is needed to keep the FP rate impact of degradation relatively under control. The new OptimizeHomogAtScale tool in ribbon_test helps to find these optimal allocation overheads for different numbers of solution columns. And Ribbon widths, with 128-bit Ribbon apparently cutting space overheads in half vs. 64-bit.

Other misc item specifics:
* Ribbon APIs in util/ribbon_config.h now provide configuration data for not just 5% construction failure rate (95% success), but also 50% and 0.1%.
  * Note that the Ribbon structure does not exhibit "threshold" behavior as standard Xor filter does, so there is a roughly fixed space penalty to cut construction failure rate in half. Thus, there isn't really an "almost sure" setting.
  * Although we can extrapolate settings for large filters, we don't have a good formula for configuring smaller filters (< 2^17 slots or so), and efforts to summarize with a formula have failed. Thus, small data is hard-coded from updated FindOccupancy tool.
* Enhances ApproximateNumEntries for public API Ribbon using more precise data (new API GetNumToAdd), thus a more accurate but not perfect reversal of CalculateSpace. (bloom_test updated to expect the greater precision)
* Move EndianSwapValue from coding.h to coding_lean.h to keep Ribbon code easily transferable from RocksDB
* Add some missing 'const' to member functions
* Small optimization to 128-bit BitParity
* Small refactoring of BandingStorage in ribbon_alg.h to support Homogeneous Ribbon
* CompactnessAndBacktrackAndFpRate now has an "expand" test: on construction failure, a possible alternative to re-seeding hash functions is simply to increase the number of slots (allocated space overhead) and try again with essentially the same hash values. (Start locations will be different roundings of the same scaled hash values--because fastrange not mod.) This seems to be as effective or more effective than re-seeding, as long as we increase the number of slots (m) by roughly m += m/w where w is the Ribbon width. This way, there is effectively an expansion by one slot for each ribbon-width window in the banding. (This approach assumes that getting "bad data" from your hash function is as unlikely as it naturally should be, e.g. no adversary.)
* 32-bit and 16-bit Ribbon configurations are added to ribbon_test for understanding their behavior, e.g. with FindOccupancy. They are not considered useful at this time and not tested with CompactnessAndBacktrackAndFpRate.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7879

Test Plan: unit test updates included

Reviewed By: jay-zhuang

Differential Revision: D26371245

Pulled By: pdillinger

fbshipit-source-id: da6600d90a3785b99ad17a88b2a3027710b4ea3a

Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/8002

Reviewed By: anand1976

Differential Revision: D26659354

Pulled By: riversand963

fbshipit-source-id: 6b464dbea9fd8240ead8cc5af393f0b78e8f9dd1

Summary:
Allow applications to implement a custom compaction filter and pass it to BlobDB.

The compaction filter's custom logic can operate on blobs.
To do so, application needs to subclass `CompactionFilter` abstract class and implement `FilterV2()` method.
Optionally, a method called `ShouldFilterBlobByKey()` can be implemented if application's custom logic rely solely
on the key to make a decision without reading the blob, thus saving extra IO. Examples can be found in
db/blob/db_blob_compaction_test.cc.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7974

Test Plan: make check

Reviewed By: ltamasi

Differential Revision: D26509280

Pulled By: riversand963

fbshipit-source-id: 59f9ae5614c4359de32f4f2b16684193cc537b39

Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/7995

Reviewed By: ltamasi

Differential Revision: D26625766

Pulled By: akankshamahajan15

fbshipit-source-id: d83c9e77695f4193da979b1ce7103b43bc1dd46c

Summary:
In the adapter class `WritableFileStringStreamAdapter`, which wraps WritableFile to be used for std::ostream, previouly only `std::endl` is considered a special case because `endl` is written by `os.put()` directly without going through `xsputn()`. `os.put()` will call `sputc()` and if we further check the internal implementation of `sputc()`, we will see it is
```
int_type __CLR_OR_THIS_CALL sputc(_Elem _Ch) {  // put a character
    return 0 < _Pnavail() ? _Traits::to_int_type(*_Pninc() = _Ch) : overflow(_Traits::to_int_type(_Ch));
```
As we explicitly disabled buffering, _Pnavail() is always 0. Thus every write, not captured by xsputn, becomes an overflow.

When I run tests on Windows, I found not only `std::endl` will drop into this case, writing an unsigned long long will also call `os.put()` then followed by `sputc()` and eventually call `overflow()`. Therefore, instead of only checking `std::endl`, we should try to append other characters as well unless the appending operation fails.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7991

Reviewed By: jay-zhuang

Differential Revision: D26615692

Pulled By: ajkr

fbshipit-source-id: 4c0003de1645b9531545b23df69b000e07014468

Summary:
RocksDB does auto-readahead for iterators on noticing more
than two reads for a table file. The readahead starts at 8KB and doubles on every
additional read upto BlockBasedTable::kMaxAutoReadAheadSize which is
256*1024.
This PR adds a new option BlockBasedTableOptions::max_auto_readahead_size which
replaces BlockBasedTable::kMaxAutoReadAheadSize and the new option can be
configured.
If max_auto_readahead_size is set 0 then no implicit auto prefetching will
be done. If max_auto_readahead_size provided is less than
8KB (which is initial readahead size used by rocksdb in case of
auto-readahead), readahead size will remain same as max_auto_readahead_size.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7951

Test Plan: Add new unit test case.

Reviewed By: anand1976

Differential Revision: D26568085

Pulled By: akankshamahajan15

fbshipit-source-id: b6543520fc74e97d859f2002328d4c5254d417af