Implement Java API for ConcurrentTaskLimiter class and compaction_thread_limiter field in ColumnFamilyOptions (#7347)

Summary:
as title

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7347

Test Plan: unit tests included

Reviewed By: jay-zhuang

Differential Revision: D23592552

Pulled By: pdillinger

fbshipit-source-id: 1c3571b6f42bfd0cfd723ff49d01fbc02a1be45b

Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/7321

Reviewed By: ajkr

Differential Revision: D23590160

fbshipit-source-id: 35d106e732ac37f674222759cdb1dbb31e005ca7

Summary:
(a) Missed a case in updating handling of rand_keys
(b) Only opening restored db with DB::Open so don't (yet)
attempt to open restored BlobDB or TransactionDB.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7361

Test Plan: better than being broken

Reviewed By: ajkr

Differential Revision: D23592570

Pulled By: pdillinger

fbshipit-source-id: dd1d999bcc0c852ee77cb6041964ec4abc0fd4fd

Fix the handling of the case when a blob file with a lower number gets added in VersionBuilder (#7349)

Summary:
When multiple background jobs are generating blob files in parallel, it is actually
possible for a blob file to be added with a file number that is lower than the
highest one in the base version. (This is a harmless race condition.) The patch
fixes the handling of this case and adds a unit test.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7349

Test Plan: `make check`

Reviewed By: riversand963

Differential Revision: D23542453

Pulled By: ltamasi

fbshipit-source-id: 4ff6f3654bc58c391d10b9870e1cc40b5e3fa8e4

Summary:
Was broken by https://github.com/facebook/rocksdb/issues/6660

Travis times before this change, after 6660:
platform_dependent: 17 min
group 1: 15 min
group 2: 44 min (often timeout on non-x86 or non-Linux)
group 3: 31 min
group 4: 21 min

After this change:
TODO

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7360

Test Plan: CI inspection

Reviewed By: ajkr

Differential Revision: D23586917

Pulled By: pdillinger

fbshipit-source-id: 4c67cf33180b0b833c39a817e6c1f128727941d2

Summary:
1.  Failed to compile because of use of FileSystem* instead of Env* to some methods;

2.  Failed to compile with addition of ConfigOptions to some methods

3.  Failed to run successfully because the database and/or db_bench would change some of the options, invalidating the comparison

4.  Failed to run successfully if Snappy was not available.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7344

Reviewed By: siying

Differential Revision: D23501093

Pulled By: jay-zhuang

fbshipit-source-id: 81fd947e95fff9db8a4c5ff419d69d4c36bef23f

Summary:
To fix the cmake build with third_party libs, like:
`mkdir build && cd build && cmake .. -DWITH_SNAPPY=1 && make`

Error:
```
Undefined symbols for architecture x86_64:
  "snappy::RawCompress(char const*, unsigned long, char*, unsigned long*)"
```

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7351

Reviewed By: pdillinger

Differential Revision: D23553705

Pulled By: jay-zhuang

fbshipit-source-id: 19b45c6763c7256107583e8af4c01d370ca06128

Summary:
Also enables a pull request to trigger all the Travis
configurations by writing FULL_CI in the commit message. (See what I did
there?)

First issue

    make: *** No rule to make target 'jl/util/crc32c_ppc_asm.o', needed by 'rocksdbjava'.  Stop.

Second issue

    tools/db_bench_tool.cc:5514:38: error: ‘gen_exp.rocksdb::Benchmark::GenerateTwoTermExpKeys::keyrange_size_’ may be used uninitialized in this function

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7359

Test Plan: CI

Reviewed By: zhichao-cao

Differential Revision: D23582132

Pulled By: pdillinger

fbshipit-source-id: 06d794673fd522ba11cf6398385387e6bd97ef89

Summary:
Replace FSRandomRWFile pointer with FSRandomRWFilePtr object in the rocksdb internal code.
This new object wraps FSRandomRWFile pointer.

Objective: If tracing is enabled, FSRandomRWFile object returns FSRandomRWFileTracingWrapper pointer that includes all necessary information in IORecord and calls underlying FileSystem and invokes IOTracer to dump that record in a binary file. If tracing is disabled then, underlying FileSystem pointer is returned directly.
FSRandomRWFilePtr wrapper class is added to bypass the FSRandomRWFileWrapper when
tracing is disabled.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7198

Test Plan: make check -j64

Reviewed By: anand1976

Differential Revision: D23421116

Pulled By: akankshamahajan15

fbshipit-source-id: 8a5ba0e7d9c1ba34c3a6f29829b107c5f09ab6a3

Summary:
gcc-4.8 returns error when using the constructor. Not sure if it's a compiler bug/limitation or code issue:
```
table/block_based/block_based_table_reader.cc:3183:67: error: use of deleted function ‘rocksdb::WritableFileStringStreamAdapter::WritableFileStringStreamAdapter(rocksdb::WritableFileStringStreamAdapter&&)’
```

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7358

Reviewed By: pdillinger

Differential Revision: D23577651

Pulled By: jay-zhuang

fbshipit-source-id: b0197e3d3538da61a6f3866410d88d2047fb9695

Summary:
As title.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7355

Reviewed By: pdillinger

Differential Revision: D23566635

Pulled By: riversand963

fbshipit-source-id: f8d846bcff637e7617b764b7bfb9a948ea18d195

Summary:
Replace FSWritableFile pointer with FSWritableFilePtr
    object in WritableFileWriter.
    This new object wraps FSWritableFile pointer.

    Objective: If tracing is enabled, FSWritableFile Ptr returns
    FSWritableFileTracingWrapper pointer that includes all necessary
    information in IORecord and calls underlying FileSystem and invokes
    IOTracer to dump that record in a binary file. If tracing is disabled
    then, underlying FileSystem pointer is returned directly.
    FSWritableFilePtr wrapper class is added to bypass the
    FSWritableFileWrapper when
    tracing is disabled.

    Test Plan: make check -j64

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7193

Reviewed By: anand1976

Differential Revision: D23355915

Pulled By: akankshamahajan15

fbshipit-source-id: e62a27a13c1fd77e36a6dbafc7006d969bed25cf

Summary:
(1) Skip check on specific key if restoring an old backup
(small minority of cases) because it can fail in those cases. (2) Remove
an old assertion about number of column families and number of keys
passed in, which is broken by atomic flush (cf_consistency) test. Like
other code (for better or worse) assume a single key and iterate over
column families. (3) Apply mock_direct_io to NewSequentialFile so that
db_stress backup works on /dev/shm.

Also add more context to output in case of backup/restore db_stress
failure.

Also a minor fix to BackupEngine to report first failure status in
creating new backup, and drop another clue about the potential
source of a "Backup failed" status.

Reverts "Disable backup/restore stress test (https://github.com/facebook/rocksdb/issues/7350)"

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7357

Test Plan:
Using backup_one_in=10000,
"USE_CLANG=1 make crash_test_with_atomic_flush" for 30+ minutes
"USE_CLANG=1 make blackbox_crash_test" for 30+ minutes
And with use_direct_reads with TEST_TMPDIR=/dev/shm/rocksdb

Reviewed By: riversand963

Differential Revision: D23567244

Pulled By: pdillinger

fbshipit-source-id: e77171c2e8394d173917e36898c02dead1c40b77

Summary:
The patch cleans up a few things in `CompactionJob::SubcompactionState`:

* Instead of using both the member initializer list and in-class initializers (and
sometimes both at the same time for the same member), the struct now uniformly
uses the latter to initialize integer members.
* The default parameter value for the constructor parameter `size` is removed.
* The explicitly deleted copy operations are removed, since they are implicitly deleted
anyways because of the `unique_ptr` members.
* The handwritten move operations, which did not move the member `c_iter` and
were not declared `nothrow`, are removed. Note that with the user-declared copy
operations gone (see the previous item), we can rely on the compiler to (correctly)
generate these methods.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7322

Test Plan: `make check`

Reviewed By: siying

Differential Revision: D23382408

Pulled By: ltamasi

fbshipit-source-id: a4ae5af150161c50ff7bdc07fa145482d0150bfe

Summary:
Currently, application may pass a statistics object to db but later
wants to reduce stats tracking overhead by setting stats level to
kExceptHistogramOrTimers (the current lowest level). Tickers will still
be incremented, causing up to 1% CPU. We can add a new lowest stats
level `kExceptTickers` to disable ticker incrementing as well, thus
reducing CPU cycles spent on tickers.

Test Plan (devserver):
```
make check
make clean
DEBUG_LEVEL=0 make db_bench
./db_bench -perf_level=1 -stats_level=0 -statistics -benchmarks=fillseq,readrandom -duration=120
```

Measure CPU util (%) before and after change:
CPU util by rocksdb::RecordTick: 1.1 vs (<0.1)

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7329

Reviewed By: pdillinger

Differential Revision: D23434014

Pulled By: riversand963

fbshipit-source-id: 72ff0f02a192ac476d4b0044b9f37fd4a22ff0d4

Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/7315

Test Plan:
`ASSERT_STATUS_CHECKED=1 make sst_dump_test && ./sst_dump_test`
And manually run `./sst_dump --file=*.sst` before and after the change.

Reviewed By: pdillinger

Differential Revision: D23361669

Pulled By: jay-zhuang

fbshipit-source-id: 5bf51a2a90ee35c8c679e5f604732ec2aef5949a

Summary:
Seems it's causing some tests failures.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7350

Reviewed By: riversand963

Differential Revision: D23544109

Pulled By: jay-zhuang

fbshipit-source-id: 798a0ca374a20b6c2d0f29582729ff101c6a2e99

Summary:
This change has the crash test randomly select from a few file
checksum implementations, or nullptr, for DB file_checksum_gen_factory.
For compatibility across runs on same DB, each non-null factory can
understand all the other functions, but the default changes.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7343

Test Plan:
'make blackbox_crash_test' for a while, including with some
debug output to ensure code is being exercised.

Reviewed By: zhichao-cao

Differential Revision: D23494580

Pulled By: pdillinger

fbshipit-source-id: 73bbc7ca32c1adaf619134c0c830f12894880b8a

Summary:
The level args should be output level instead of input levels.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7346

Test Plan: make check

Reviewed By: ajkr

Differential Revision: D23506373

Pulled By: cheng-chang

fbshipit-source-id: b2f701d44c13581c5c10c4dbebded4fcd354d641

Summary:
Although added to db_stress, testing of backup/restore
was never integrated into the crash test, originally concerned about
performance. I've enabled it now and to address the peformance concern,
testing backup/restore is always skipped once the db exceeds a certain
size threshold, default 100MB. This should provide sufficient
opportunity for testing BackupEngine without bogging down everything
else with heavier and heavier operations.

Also fixed backup/restore in db_stress by making sure PurgeOldBackups
can remove manifest files, which are normally kept around for db_stress.

Added more coverage of backup options, and up to three backups being
saved in one backup directory (in some cases).

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7348

Test Plan:
ran 'make blackbox_crash_test' for a while, with heightened
probabilitly of taking backups (1/10k). Also confirmed with some debug
output that the code is being covered, TestBackupRestore only takes
a few seconds to complete when triggered, and even at 1/10k and ~50MB
database, there's <,~ 1 thread testing backups at any time.

Reviewed By: ajkr

Differential Revision: D23510835

Pulled By: pdillinger

fbshipit-source-id: b6b8735591808141f81f10773ac31634cf03b6c0

Summary:
This is adapted from https://github.com/facebook/rocksdb/issues/6678 but takes a different approach, avoiding opening a read-write DB and avoiding the `DeleteFile()` API.

First, this PR refactors how options variables are initialized in `ldb` so it can be reused in a subcommand that doesn't open a DB:

- Separated remaining option initialization logic out of `OpenDB()`. The new `PrepareOptions()` function initializes the full options state.
- Fixed an old TODO about applying the subcommand CF option overrides to the proper `ColumnFamilyOptions` object.

Second, this PR adds the `ldb unsafe_remove_sst_file` subcommand. It uses the `VersionSet`-level APIs to remove the file with the specified number.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7335

Test Plan: played with interactive python and this file removal command. Verified openability/correct results in case of multiple column families, multiple levels, etc.

Reviewed By: pdillinger

Differential Revision: D23454575

Pulled By: ajkr

fbshipit-source-id: 039b7a8cbfc42fd123dcb25821eef51d61148afe

Summary:
Since we can't land https://github.com/facebook/rocksdb/issues/7336 until the next major release, added a strong warning against the `DeleteFile()` API in the meantime.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7337

Reviewed By: pdillinger

Differential Revision: D23459728

Pulled By: ajkr

fbshipit-source-id: 326cb9b18190386080c35c761a8736d8a877dafb

Summary:
In block-based table builder, the cut-over from buffered to unbuffered
mode involves sampling the buffered blocks and generating a dictionary.
There was a bug where `SstFileWriter` passed zero as the `target_file_size`
causing the cutover to happen immediately, so there were no samples
available for generating the dictionary.

This PR changes the meaning of `target_file_size == 0` to mean buffer
the whole file before cutting over. It also adds dictionary compression
support to `sst_dump --command=recompress` for easy evaluation.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7323

Reviewed By: cheng-chang

Differential Revision: D23412158

Pulled By: ajkr

fbshipit-source-id: 3b232050e70ef3c2ee85a4b5f6fadb139c569873

Summary:
This PR creates `rocksdb_open_column_families_with_ttl` which allows C API users to open a DBWithTLL with column families.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7314

Reviewed By: cheng-chang

Differential Revision: D23430287

Pulled By: ajkr

fbshipit-source-id: 307aa21d170d1402653263a91f6f832ef76afba0

Summary:
- Closes https://github.com/facebook/rocksdb/issues/6490
- Currently MERGEs are converted to PUTs at bottom or compaction has reached the beginning of the key, this can wrongly cover a PUT future base case.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7166

Test Plan:
- Automated: `make all check`
- Manual: With `allow_ingest_behind = true`, add Merge operations to a key then run compaction. Then run ingesting external files to make sure the base case is probably compacted with existing Merges.

Reviewed By: cheng-chang

Differential Revision: D23325425

Pulled By: ajkr

fbshipit-source-id: 3eb415eb7b381b5453e45245393566153b1abb68

Summary:
Delete database instances to make sure there are no loose threads
running before exit(). This fixes segfaults seen when running
workloads through CompositeEnvs with custom file systems.

For further background on the issues arising when using CompositeEnvs, see the discussion in:
https://github.com/facebook/rocksdb/pull/6878

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7327

Reviewed By: cheng-chang

Differential Revision: D23433244

Pulled By: ajkr

fbshipit-source-id: 4e19cf2067e3fe68c2a3fe1823f24b4091336bbe

Summary:
These new functions and 128-bit value bit operations are
expected to be used in a forthcoming Bloom filter alternative.

No functional changes to production code, just new code only called by
unit tests, cosmetic changes to existing headers, and fix an existing
function for a yet-unused template instantiation (BitsSetToOne on
something signed and smaller than 32 bits).

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7338

Test Plan:
Unit tests included. Works with and without
TEST_UINT128_COMPAT=1 to check compatibility with and without
__uint128_t. Also added that parameter to the CircleCI build
build-linux-shared_lib-alt_namespace-status_checked.

Reviewed By: jay-zhuang

Differential Revision: D23494945

Pulled By: pdillinger

fbshipit-source-id: 5c0dc419100d9df5d4d9abb153b2855d5aea39e8

Summary:
This PR is set up to merge into master, but it would be great to get this into a patch release if possible.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7334

Reviewed By: jay-zhuang

Differential Revision: D23476624

Pulled By: pdillinger

fbshipit-source-id: c6cc02ce06e779e1e174ab0f4748e557d2ce7bc6

Summary:
Fix typo in comment for SeekForGetImpl().
Rename "bounary" to "boundary"

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7328

Reviewed By: riversand963

Differential Revision: D23439748

Pulled By: zhichao-cao

fbshipit-source-id: 83a34c417c71a3210ce54a090d76c4d5571313f3

Summary:
L0 score is based on size target and number of files. The size target
used is `max_bytes_for_level_base`. However, the base level's size can
dynamically expand in write burst mode. In fact, it can expand so much
that L0->Lbase becomes the highest fanout in target sizes. This doesn't
make sense from an efficiency perspective, so this PR bounds the
L0->Lbase fanout to the smoothed level multiplier. The L0 scoring based
on file count remains unchanged.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7325

Test Plan:
contrived benchmark that exhibits the problem:

```
$ TEST_TMPDIR=/data/users/andrewkr/ ./db_bench -benchmarks=filluniquerandom,readrandom -write_buffer_size=1048576 -target_file_size_base=1048576 -max_bytes_for_level_base=4194304 -level0_file_num_compaction_trigger=4 -level_compaction_dynamic_level_bytes=true -compression_type=none -max_background_jobs=12 -rate_limiter_bytes_per_sec=104857600 -benchmark_write_rate_limit=10485760 -num=100000000
```

Results:

- "Burst W-Amp" is the write-amp near the end of the fillrandom benchmark
- "Total W-Amp" is the write-amp after readrandom has run a while and all levels no longer need compaction

Branch | Burst W-Amp | Total W-Amp | fillrandom (MB/s)
-- | -- | -- | --
master | 20.2 | 21.5 | 4.7
dynamic-l0-score | 12.6 | 14.1 | 7.2

Reviewed By: siying

Differential Revision: D23412935

Pulled By: ajkr

fbshipit-source-id: f91f2067188e432dd39deab02f1c56f195057a0e

Summary:
Quick fixes to examples to make it easier to get familiar with RocksDB for Windows users:

- Set proper temporary directory path on Windows for all examples (with C++17 we should start using std::filesystem)
- Fixed typo and got rid of warnings treated as errors in multi_processes_example.cc
- Get number of available cores on Windows in c_simple_example.c
- Add command to remove DB directory for Windows in compaction_filter_example.cc (print error, but carry on with example upon error, because error code is returned if there is no such directory on Windows)

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7304

Reviewed By: jay-zhuang

Differential Revision: D23450900

Pulled By: pdillinger

fbshipit-source-id: 4256134deb6ae6bb267ed1bd69f814842b95f60f

Summary:
Recompress blobs if compression type is changed.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7331

Test Plan: `make check`

Reviewed By: ltamasi

Differential Revision: D23437102

Pulled By: jay-zhuang

fbshipit-source-id: bb699ebdad137721d422e42e331d4de8a82a7c5f

Summary:
The patch adds a log message to `BlobFileBuilder` that is logged upon
generating a blob file, similarly to how we log the generation of table files
during flush and compaction. The log message contains the column family
name, job id, blob file number, and the number and total size of blobs in
the new file.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7324

Test Plan: Ran `make check` and checked the actual log messages using a custom `db_bench`.

Reviewed By: riversand963

Differential Revision: D23402229

Pulled By: ltamasi

fbshipit-source-id: ca42beb4db284b783d1eb2651f321032a45d0c5f

Summary:
Add a unit test case to check memory usage when
max_write_buffer_size_to_maintain is set if flushed immutable memtables are
trimmed timely or not.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7311

Test Plan: Compared the results with before bug fix.

Reviewed By: ltamasi

Differential Revision: D23321702

Pulled By: akankshamahajan15

fbshipit-source-id: da04ee21137d641a07fd499a9e2749eb036fcb1e

Summary:
Fixes https://github.com/facebook/rocksdb/issues/7185.

Standard for GNU and FreeBSD. See
https://www.freebsd.org/doc/en/books/porters-handbook/porting-prefix.html
https://www.gnu.org/prep/standards/html_node/DESTDIR.html#DESTDIR
https://www.gnu.org/prep/standards/html_node/Directory-Variables.html

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7187

Reviewed By: cheng-chang

Differential Revision: D23333233

Pulled By: ajkr

fbshipit-source-id: f704d23852c4516cf5fa00df73ff57687b2ddffb

Summary:
A new file interface `SupportPrefetch()` is added. When the user overrides it to `false`, an internal prefetch buffer will be used for readahead. Useful for non-directIO but FS doesn't have readahead support.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7312

Reviewed By: anand1976

Differential Revision: D23329847

Pulled By: jay-zhuang

fbshipit-source-id: 71cd4ce6f4a820840294e4e6aec111ab76175527

Summary:
The patch adds a class called `BlobFileBuilder` that can be used to build
and cut blob files in background jobs (flushes/compactions). The class
enforces a value size threshold (`min_blob_size`; smaller blobs will be inlined
in the LSM tree itself), and supports specifying a blob file size limit (`blob_file_size`),
as well as compression (`blob_compression_type`) and checksums for blob files.
It also keeps track of the generated blob files and their associated `BlobFileAddition`
metadata, which can be applied as part of the background job's `VersionEdit`.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7306

Test Plan: `make check`

Reviewed By: riversand963

Differential Revision: D23298817

Pulled By: ltamasi

fbshipit-source-id: 38f35d81dab1ba81f15236240612ec173d7f21b5

Summary:
Replace FSRandomAccessFile pointer with FSRandomAccessFilePtr
    object in RandomAccessFileReader.
    This new object wraps FSRandomAccessFile pointer.

    Objective: If tracing is enabled, FSRandomAccessFile Ptr returns
    FSRandomAccessFileTracingWrapper pointer that includes all necessary
    information in IORecord and calls underlying FileSystem and invokes
    IOTracer to dump that record in a binary file. If tracing is disabled
    then, underlying FileSystem pointer is returned directly.
    FSRandomAccessFilePtr wrapper class is added to bypass the FSRandomAccessFileWrapper when
    tracing is disabled.

    Test Plan: make check -j64

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7192

Reviewed By: anand1976

Differential Revision: D23356867

Pulled By: akankshamahajan15

fbshipit-source-id: 48f31168166a17a7444b40be44a9a9d4a5c7182c

Summary:
This is a "real" fix for the issue worked around in https://github.com/facebook/rocksdb/issues/7294.
To get DB checksum info for live files, we now read the manifest file
that will become part of the checkpoint/backup. This requires a little
extra handling in taking a custom checkpoint, including only reading the
manifest file up to the size prescribed by the checkpoint.

This moves GetFileChecksumsFromManifest from backup code to
file_checksum_helper.{h,cc} and removes apparently unnecessary checking
related to column families.

Updated HISTORY.md and warned potential future users of
DB::GetLiveFilesChecksumInfo()

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7309

Test Plan: updated unit test, before and after

Reviewed By: ajkr

Differential Revision: D23311994

Pulled By: pdillinger

fbshipit-source-id: 741e30a2dc1830e8208f7648fcc8c5f000d4e2d5

Summary:
Right now all I/O failures under PartitionIndexReader::CacheDependencies() is swallowed. This doesn't impact correctness but we've made a decision that any I/O error in read path now should be returned to users for awareness. Return errors in those cases instead.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7297

Test Plan: Add a new unit test that ingest errors in this code path and see Get() fails. Only one I/O path is hit in PartitionIndexReader::CacheDependencies(). Several option changes are attempt but not able to got other pread paths triggered. Not sure whether other failure cases would be even possible. Would rely on continuous stress test to validate it.

Reviewed By: anand1976

Differential Revision: D23257950

fbshipit-source-id: 859dbc92fa239996e1bb378329344d3d54168c03