Summary:
Add ZenFS, a file system for zoned block devices, to PLUGINS.md

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8218

Reviewed By: jay-zhuang

Differential Revision: D27944376

Pulled By: ajkr

fbshipit-source-id: c9ea2e9814001ccd7c56d7ef4d38e20dfeb48d1e

Summary:
In current RocksDB, in recover the information form WAL, we do the consistency check for each column family when one WAL file is corrupted and PointInTimeRecovery is set. However, it will report a false positive alert on "SST file is ahead of WALs" when one of the CF current log number is greater than the corrupted WAL number (CF contains the data beyond the corrupted WAl) due to a new column family creation during flush. In this case, a new WAL is created (it is empty) during a flush. Also, due to some reason (e.g., storage issue or crash happens before SyncCloseLog is called), the old WAL is corrupted. The new CF has no data, therefore, it does not have the consistency issue.

Fix: when checking cfd->GetLogNumber() > corrupted_wal_number also check cfd->GetLiveSstFilesSize() > 0. So the CFs with no SST file data will skip the check here.

Note potential ignored inconsistency caused due to fix: empty CF can also be caused by write+delete. In this case, after flush, there is no SST files being generated. However, this CF still have the log in the WAL. When the WAL is corrupted, the DB might be inconsistent.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8207

Test Plan: added unit test, make crash_test

Reviewed By: riversand963

Differential Revision: D27898839

Pulled By: zhichao-cao

fbshipit-source-id: 931fc2d8b92dd00b4169bf84b94e712fd688a83e

Summary:
For some compilers/environments (e.g. Clang, riscv64), we need to link against -latomic.  Check if this is a requirement and add the library to the third-party libs if it is.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8183

Reviewed By: pdillinger

Differential Revision: D27773564

Pulled By: mrambacher

fbshipit-source-id: 68e15d823144f83fb02221c7bf5b1e43323419bf

Summary:
RocksDB allows user-specified custom comparators which may not be known to `ldb`,
a built-in tool for checking/mutating the database. Therefore, column family comparator
names mismatch encountered during manifest dump should not prevent the dumping from
proceeding.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8216

Test Plan:
```
make check
```

Also manually do the following
```
KEEP_DB=1 ./db_with_timestamp_basic_test
./ldb --db=<db> manifest_dump --verbose
```
The ldb should succeed and print something like:
```
...
--------------- Column family "default"  (ID 0) --------------
log number: 6
comparator: <TestComparator>, but the comparator object is not available.
...
```

Reviewed By: ltamasi

Differential Revision: D27927581

Pulled By: riversand963

fbshipit-source-id: f610b2c842187d17f575362070209ee6b74ec6d4

Summary:
Add comment to DisableManualCompaction() which was missing.
Also explictly return from DBImpl::CompactRange() to avoid memtable flush when manual compaction is disabled.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8186

Test Plan: Run existing unit tests.

Reviewed By: jay-zhuang

Differential Revision: D27744517

fbshipit-source-id: 449548a48905903b888dc9612bd17480f6596a71

Summary:
When WriteBufferManager is shared across DBs and column families
to maintain memory usage under a limit, OOMs have been observed when flush cannot
finish but writes continuously insert to memtables.
In order to avoid OOMs, when memory usage goes beyond buffer_limit_ and DBs tries to write,
this change will stall incoming writers until flush is completed and memory_usage
drops.

Design: Stall condition: When total memory usage exceeds WriteBufferManager::buffer_size_
(memory_usage() >= buffer_size_) WriterBufferManager::ShouldStall() returns true.

DBImpl first block incoming/future writers by calling write_thread_.BeginWriteStall()
(which adds dummy stall object to the writer's queue).
Then DB is blocked on a state State::Blocked (current write doesn't go
through). WBStallInterface object maintained by every DB instance is added to the queue of
WriteBufferManager.

If multiple DBs tries to write during this stall, they will also be
blocked when check WriteBufferManager::ShouldStall() returns true.

End Stall condition: When flush is finished and memory usage goes down, stall will end only if memory
waiting to be flushed is less than buffer_size/2. This lower limit will give time for flush
to complete and avoid continous stalling if memory usage remains close to buffer_size.

WriterBufferManager::EndWriteStall() is called,
which removes all instances from its queue and signal them to continue.
Their state is changed to State::Running and they are unblocked. DBImpl
then signal all incoming writers of that DB to continue by calling
write_thread_.EndWriteStall() (which removes dummy stall object from the
queue).

DB instance creates WBMStallInterface which is an interface to block and
signal DBs during stall.
When DB needs to be blocked or signalled by WriteBufferManager,
state_for_wbm_ state is changed accordingly (RUNNING or BLOCKED).

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7898

Test Plan: Added a new test db/db_write_buffer_manager_test.cc

Reviewed By: anand1976

Differential Revision: D26093227

Pulled By: akankshamahajan15

fbshipit-source-id: 2bbd982a3fb7033f6de6153aa92a221249861aae

Summary:
This partially reverts commit 10196d7e.

The problem with this change is because of important filter use cases:
FIFO compaction and SST writer. FIFO "compaction" always uses level 0 so
would only use Ribbon filters if specifically including level 0 for the
Ribbon filter policy. SST writer sets level_at_creation=-1 to indicate
unknown level, and this would be treated the same as level 0 unless
fixed.

We are keeping the part about committing to permanent schema, which is
only changes to API comments and HISTORY.md.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8212

Test Plan: CI

Reviewed By: jay-zhuang

Differential Revision: D27896468

Pulled By: pdillinger

fbshipit-source-id: 50a775f7cba5d64fb729d9b982e355864020596e

Summary:
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8208

Make include of "file_system.h" use the same include path as everywhere
else.

Reviewed By: riversand963, akankshamahajan15

Differential Revision: D27881606

fbshipit-source-id: fc1e076229fde21041a813c655ce017b5070c8b3

Summary:
Fixes https://github.com/facebook/rocksdb/issues/6245.

Adapted from https://github.com/facebook/rocksdb/issues/8201 and https://github.com/facebook/rocksdb/issues/8205.

Previously we were writing the ingested file's smallest/largest internal keys
with sequence number zero, or `kMaxSequenceNumber` in case of range
tombstone. The former (sequence number zero) is incorrect and can lead
to files being incorrectly ordered. The fix in this PR is to overwrite
boundary keys that have sequence number zero with the ingested file's assigned
sequence number.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8209

Test Plan: repro unit test

Reviewed By: riversand963

Differential Revision: D27885678

Pulled By: ajkr

fbshipit-source-id: 4a9f2c6efdfff81c3a9923e915ea88b250ee7b6a

Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/8210

Reviewed By: akankshamahajan15

Differential Revision: D27887612

Pulled By: ltamasi

fbshipit-source-id: 0db8d0b6047334dc47fe30a98804449043454386

Summary:
Unittest reports no space from time to time, which can be reproduced on a small memory machine with SHM. It's caused by large WAL files generated during the test, which is preallocated, but didn't truncate during close(). Adding the missing APIs to set preallocation.
It added arm test as nightly build, as the test runs more than 1 hour.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8204

Test Plan: test on small memory arm machine

Reviewed By: mrambacher

Differential Revision: D27873145

Pulled By: jay-zhuang

fbshipit-source-id: f797c429d6bc13cbcc673bc03fcc72adda55f506

Summary:
Moving ARM build from travis to CircleCI.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8203

Test Plan: CI

Reviewed By: ajkr

Differential Revision: D27861753

Pulled By: jay-zhuang

fbshipit-source-id: 5e36a67f6fbb921c2ed80b284ba2de485411937b

Summary:
In a distributed environment, a file `rename()` operation can succeed on server (remote)
side, but the client can somehow return non-ok status to RocksDB. Possible reasons include
network partition, connection issue, etc. This happens in `rocksdb::SetCurrentFile()`, which
can be called in `LogAndApply() -> ProcessManifestWrites()` if RocksDB tries to switch to a
new MANIFEST. We currently always delete the new MANIFEST if an error occurs.

This is problematic in distributed world. If the server-side successfully updates the CURRENT
file via renaming, then a subsequent `DB::Open()` will try to look for the new MANIFEST and fail.

As a fix, we can track the execution result of IO operations on the new MANIFEST.
- If IO operations on the new MANIFEST fail, then we know the CURRENT must point to the original
  MANIFEST. Therefore, it is safe to remove the new MANIFEST.
- If IO operations on the new MANIFEST all succeed, but somehow we end up in the clean up
  code block, then we do not know whether CURRENT points to the new or old MANIFEST. (For local
  POSIX-compliant FS, it should still point to old MANIFEST, but it does not matter if we keep the
  new MANIFEST.) Therefore, we keep the new MANIFEST.
    - Any future `LogAndApply()` will switch to a new MANIFEST and update CURRENT.
    - If process reopens the db immediately after the failure, then the CURRENT file can point
      to either the new MANIFEST or the old one, both of which exist. Therefore, recovery can
      succeed and ignore the other.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8192

Test Plan: make check

Reviewed By: zhichao-cao

Differential Revision: D27804648

Pulled By: riversand963

fbshipit-source-id: 9c16f2a5ce41bc6aadf085e48449b19ede8423e4

Summary:
Historically, the DB properties `rocksdb.cur-size-active-mem-table`,
`rocksdb.cur-size-all-mem-tables`, and `rocksdb.size-all-mem-tables` called
the method `MemTable::ApproximateMemoryUsage` for mutable memtables,
which is not safe without synchronization. This resulted in data races with
memtable inserts. The patch changes the code handling these properties
to use `MemTable::ApproximateMemoryUsageFast` instead, which returns a
cached value backed by an atomic variable. Two test cases had to be updated
for this change. `MemoryTest.MemTableAndTableReadersTotal` was fixed by
increasing the value size used so each value ends up in its own memtable,
which was the original intention (note: the test has been broken in the sense
that the test code didn't consider that memtable sizes below 64 KB get
increased to 64 KB by `SanitizeOptions`, and has been passing only by
accident). `DBTest.MemoryUsageWithMaxWriteBufferSizeToMaintain` relies on
completely up-to-date values and thus was changed to use `ApproximateMemoryUsage`
directly instead of going through the DB properties. Note: this should be safe in this case
since there's only a single thread involved.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8206

Test Plan: `make check`

Reviewed By: riversand963

Differential Revision: D27866811

Pulled By: ltamasi

fbshipit-source-id: 7bd754d0565e0a65f1f7f0e78ffc093beef79394

Summary:
If `options.best_efforts_recovery == true`, RocksDB currently tolerates missing table files and recovers to the latest version without missing table files (not considering WAL). It is necessary to handle blob files as well to make the feature more complete.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8180

Test Plan: make check

Reviewed By: ltamasi

Differential Revision: D27840556

Pulled By: riversand963

fbshipit-source-id: 041685d0dc2e7779ac4f0374c07a8a327704aa5e

Summary:
Test was flaky because for kUseDbSessionId naming, blob files use
naming scheme kLegacyCrc32cAndFileSize. So expected number of files
because of collision can vary. So disabling blobdb for this test case.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8197

Reviewed By: pdillinger

Differential Revision: D27836997

Pulled By: akankshamahajan15

fbshipit-source-id: 5eb21a5f4acae3d6b730a9e1b207264fbc18cb80

Summary:
Update release version to 6.20

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8199

Test Plan: No code change

Reviewed By: ajkr

Differential Revision: D27838750

Pulled By: akankshamahajan15

fbshipit-source-id: f02f722fc6bdd37d626d47a0e932bbecea3507a8

Summary:
Since the Ribbon filter schema seems good (compatible back to
6.15.0), this change commits to long term support of the SST schema,
even though we expect the API for enabling Ribbon to change (still
called NewExperimentalRibbonFilterPolicy).

This also adds support for "hybrid" configuration in which some levels
use Bloom (higher levels, lower numbered) for speed and the rest use
Ribbon (lower levels, higher numbered) for memory space efficiency.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8198

Test Plan: unit test added, crash test support

Reviewed By: jay-zhuang

Differential Revision: D27831232

Pulled By: pdillinger

fbshipit-source-id: 90e528677689474d293ed6710b42ba89fbd5b5ab

Summary:
The code for strcmp that was present does work when compiled for Windows unicode file paths.

Needs backporting to:
* 6.17.fb
* 6.18.fb
* 6.19.fb

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8190

Reviewed By: akankshamahajan15

Differential Revision: D27765588

Pulled By: jay-zhuang

fbshipit-source-id: 89f8a5ac61fd7edc758340dfd335b0a5f96dae6e

Summary:
- Fixes the makefile to do the right thing when invoking multiple targets (e.g. make shared_lib install-shared).

- Fixes the building of db_stress in shared lib mode.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8195

Reviewed By: pdillinger

Differential Revision: D27803452

Pulled By: mrambacher

fbshipit-source-id: 7c285d267770a359eb47f25855affdf58687e0e4

Summary:
Added the Blob option settings from the AdvancedColmnFamilyOptions to the C API.

There are no tests for getting/setting options in the C API currently, hence no specific test plans.  Should there be a some?

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8148

Reviewed By: ltamasi

Differential Revision: D27568495

Pulled By: mrambacher

fbshipit-source-id: 3a52b784467ea2c4bc58be5f75c5d41f0a5c55d6

Summary:
Updated the test to wait until all trash files are deleted by
SSTFileManager in the background. Since deletion runs in background so
number of files deleted might not always be as expected.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8196

Reviewed By: jay-zhuang

Differential Revision: D27812273

Pulled By: akankshamahajan15

fbshipit-source-id: d3ace1db34f91254b52fa455e09844d02801f58e

Summary:
Fix for failure to build in LITE mode on MacOs from
BlobFileCompletionCallback unused private fields.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8189

Reviewed By: jay-zhuang

Differential Revision: D27768341

Pulled By: akankshamahajan15

fbshipit-source-id: 14d31d7a9b52d308d9f9f27feff1977c5550622f

Summary:
Extend the DB::GetLiveFilesChecksumInfo API to blob files.
This API is also used by the file_checksum_dump ldb command to dump checksum
of SST files which now also dumps blob files checksum.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8179

Test Plan: Add new unit test

Reviewed By: zhichao-cao

Differential Revision: D27714965

Pulled By: akankshamahajan15

fbshipit-source-id: d8b7343ea845a64c83800336d88cced7152a8c92

Summary:
As the name of `DBImpl::WriteLevel0TableForRecovery` suggests, the resulting table file
should be placed on L0. However, the argument `level` passed to `BuildTable()` is -1.

We need to correct this since the level information will be useful to determine file placement.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8187

Test Plan: make check

Reviewed By: ltamasi

Differential Revision: D27748570

Pulled By: riversand963

fbshipit-source-id: e1cd23128a8de31f14b1edc2ea92754c154e4f10

Summary:
Resolves https://github.com/facebook/rocksdb/issues/8014

- Add an assertion on `DB::Open` to ensure `db_options.max_open_files` is unlimited if FIFO Compaction is being used.
- This is to align with what the docs mention and to prevent premature data deletion.
- Update tests to work with this assertion.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8172

Test Plan:
```bash
$ make check -j$(nproc)

Generated TARGETS Summary:
- 6 libs
- 0 binarys
- 180 tests
```

Reviewed By: ajkr

Differential Revision: D27768792

Pulled By: thejchap

fbshipit-source-id: cf6350535e3a3577fec72bcba75b3c094dc7a6f3

Summary:
Add a blog post that calls for contribution in incremental compaction

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8182

Reviewed By: ajkr

Differential Revision: D27724150

fbshipit-source-id: 42e474858b286a53e5aaa1c4e7242a8c745af651

Summary:
Before this PR, `get_iostats_context()` will silently return a nullptr if no thread_local support is detected.
This can be the result of build_detect_platform's failure to compile the simple code snippet on certain platforms, as
reported in https://github.com/facebook/mysql-5.6/issues/904.
To be safe, we should fail the compilation if user does not opt out IOStatsContext and
ROCKSDB_SUPPORT_THREAD_LOCAL is not defined.

If RocksDB relies on c++11, can we just always use thread_local? It turns out there might be
performance concerns (https://github.com/facebook/rocksdb/issues/5774),
which is beyond the scope of this PR. We can revisit this later. Here, we stick to the original impl.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8117

Reviewed By: ajkr

Differential Revision: D27356847

Pulled By: riversand963

fbshipit-source-id: f7d5776842277598d8341b955febb601946801ae

Summary:
* CreateNewBackup(WithMetadata) returning the BackupID of new backup
through optional new output param. This is especially useful with the
new mutithreading support, so that you can transactionally determine the
ID of a backup you create.
* GetBackupInfo / GetLatestBackupInfo for individual backups, so that
you don't have to comb through a vector of backups if you don't want to.

Updated HISTORY.md (including re: BlobDB support as new feature)

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8170

Test Plan:
Added test logic to existing tests, to minimize increase in
cost of running tests

Reviewed By: zhichao-cao

Differential Revision: D27680410

Pulled By: pdillinger

fbshipit-source-id: 1fc45b73d81aae293ccd4a43d9583d7fd915d3eb

Summary:
When compiling RocksDB with Buck for ARM64, the linker complains about missing crc32 symbols that are defined in the crc32c_arm64.cc file. Since this file wasn't included in the build this is totally expected

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8168

Test Plan:
The following no longer fails to link rocksdb:
  buck build mode/mac-xcode //eden/fs/service:edenfs#macosx-arm64

Reviewed By: zhichao-cao

Differential Revision: D27664627

Pulled By: xavierd

fbshipit-source-id: fb9d7a538599ee7a08882f87628731de6e641f8d

Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/8160

Reviewed By: riversand963

Differential Revision: D27672474

Pulled By: ajkr

fbshipit-source-id: fdbbc3398f0b1d4cef6b68636e5caf369c34b3a7

Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/8169

Reviewed By: riversand963

Differential Revision: D27672457

Pulled By: ajkr

fbshipit-source-id: b40a7ad5d09a754154f28c2574ef9f77c8a131bb

Summary:
Current flush reason attribution is misleading or incorrect (depending on what the original intention was):

- Flush due to WAL reaching its maximum size is attributed to `kWriteBufferManager`
- Flushes due to full write buffer and write buffer manager are not distinguishable, both are attributed to `kWriteBufferFull`

This changes the first to a new flush reason `kWALFull`, and splits the second between `kWriteBufferManager` and `kWriteBufferFull`.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8150

Reviewed By: zhichao-cao

Differential Revision: D27569645

Pulled By: ot

fbshipit-source-id: 7e3c8ca186a6e71976e6b8e937297eebd4b769cc

Summary:
Enable backup/restore functionality with Integrated BlobDB in
db_stress and crash test.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8165

Test Plan:
Ran python3 -u tools/db_crashtest.py --simple whitebox along
with :
  1. decreased "backup_in_one" value for backups to be more frequent and
  2. manually changed code for "enable_blob_file" to be always true and
     apply blobdb params 100% for testing purpose.

Reviewed By: ltamasi

Differential Revision: D27636025

Pulled By: akankshamahajan15

fbshipit-source-id: 0d0e0d1479ced163f992872dc998e79c581bfc99

Summary:
Add support for blob files for backup/restore like table files.
    Since DB session ID is currently not supported for blob files (there is no place to store it in
    the header), so for blob files uses the
    kLegacyCrc32cAndFileSize naming scheme even if
    share_files_with_checksum_naming is set to kUseDbSessionId.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8129

Test Plan: Add new test units

Reviewed By: ltamasi

Differential Revision: D27408510

Pulled By: akankshamahajan15

fbshipit-source-id: b27434d189a639ef3e6ad165c61a143a2daaf06e

Summary:
Fixing another crash test failure in the case of
write_dbid_to_manifest=true and reading a backup as read-only DB.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8164

Test Plan:
enhanced unit test for backup as read-only DB, ran
blackbox_crash_test more with elevated backup_one_in

Reviewed By: zhichao-cao

Differential Revision: D27622237

Pulled By: pdillinger

fbshipit-source-id: 680d0f99ddb465a601737f2e3f2c80efd47384fb

Summary:
Forgot to re-test crash test after adding read-only filesystem
enforcement to https://github.com/facebook/rocksdb/issues/8142. The problem is ReadOnlyFileSystem would reject
CreateDirIfMissing whenever DBOptions::create_if_missing=true. The fix
that is better for users is to allow CreateDirIfMissing in
ReadOnlyFileSystem if the directory exists, so that they don't cause a
failure on using create_if_missing with opening backups as read-only
DBs. Added this option test to the unit test (in addition to being in the
crash test).

Also fixed a couple of lints.

And some better messaging from 'make format' so that when you run it
with uncommitted changes, it's clear that it's only checking the
uncommitted changes.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8161

Test Plan: local blackbox_crash_test with amplified backup_one_in

Reviewed By: ajkr

Differential Revision: D27614409

Pulled By: pdillinger

fbshipit-source-id: 63ccb626c7e34c200d61c6bca2a8f60da9015179

Summary:
Updated instructions for installing zstd on CentOS and note about clang-format dependency

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8158

Reviewed By: riversand963

Differential Revision: D27598665

Pulled By: ajkr

fbshipit-source-id: e349eeb91147f3163e170cc29c8460b06d739b5b

Summary:
A current limitation of backups is that you don't know the
exact database state of when the backup was taken. With this new
feature, you can at least inspect the backup's DB state without
restoring it by opening it as a read-only DB.

Rather than add something like OpenAsReadOnlyDB to the BackupEngine API,
which would inhibit opening stackable DB implementations read-only
(if/when their APIs support it), we instead provide a DB name and Env
that can be used to open as a read-only DB.

Possible follow-up work:

* Add a version of GetBackupInfo for a single backup.
* Let CreateNewBackup return the BackupID of the newly-created backup.

Implementation details:

Refactored ChrootFileSystem to split off new base class RemapFileSystem,
which allows more general remapping of files. We use this base class to
implement BackupEngineImpl::RemapSharedFileSystem.

To minimize API impact, I decided to just add these fields `name_for_open`
and `env_for_open` to those set by GetBackupInfo when
include_file_details=true. Creating the RemapSharedFileSystem adds a bit
to the memory consumption, perhaps unnecessarily in some cases, but this
has been mitigated by (a) only initialize the RemapSharedFileSystem
lazily when GetBackupInfo with include_file_details=true is called, and
(b) using the existing `shared_ptr<FileInfo>` objects to hold most of the
mapping data.

To enhance API safety, RemapSharedFileSystem is wrapped by new
ReadOnlyFileSystem which rejects any attempts to write. This uncovered a
couple of places in which DB::OpenForReadOnly would write to the
filesystem, so I fixed these. Added a release note because this affects
logging.

Additional minor refactoring in backupable_db.cc to support the new
functionality.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8142

Test Plan:
new test (run with ASAN and UBSAN), added to stress test and
ran it for a while with amplified backup_one_in

Reviewed By: ajkr

Differential Revision: D27535408

Pulled By: pdillinger

fbshipit-source-id: 04666d310aa0261ef6b2385c43ca793ce1dfd148

Summary:
According to https://github.com/facebook/rocksdb/issues/5907, each filter partition "should include the bloom of the prefix of the last
key in the previous partition" so that SeekForPrev() in prefix mode can return correct result.
The prefix of the last key in the previous partition does not necessarily have the same prefix
as the first key in the current partition. Regardless of the first key in current partition, the
prefix of the last key in the previous partition should be added. The existing code, however,
does not follow this. Furthermore, there is another issue: when finishing current filter partition,
`FullFilterBlockBuilder::AddPrefix()` is called for the first key in next filter partition, which effectively
overwrites `last_prefix_str_` prematurely. Consequently, when the filter block builder proceeds
to the next partition, `last_prefix_str_` will be the prefix of its first key, leaving no way of adding
the bloom of the prefix of the last key of the previous partition.

Prefix extractor is FixedLength.2.
```
[  filter part 1   ]    [  filter part 2    ]
                  abc    d
```
When SeekForPrev("abcd"), checking the filter partition will land on filter part 2 because "abcd" > "abc"
but smaller than "d".
If the filter in filter part 2 happens to return false for the test for "ab", then SeekForPrev("abcd") will build
incorrect iterator tree in non-total-order mode.

Also fix a unit test which starts to fail following this PR. `InDomain` should not fail due to assertion
error when checking on an arbitrary key.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8137

Test Plan:
```
make check
```

Without this fix, the following command will fail pretty soon.
```
./db_stress --acquire_snapshot_one_in=10000 --avoid_flush_during_recovery=0 \
--avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 \
--batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=17 \
--bottommost_compression_type=disable --cache_index_and_filter_blocks=1 --cache_size=1048576 \
--checkpoint_one_in=0 --checksum_type=kxxHash64 --clear_column_family_one_in=0 \
--compact_files_one_in=1000000 --compact_range_one_in=1000000 --compaction_ttl=0 \
--compression_max_dict_buffer_bytes=0 --compression_max_dict_bytes=0 \
--compression_parallel_threads=1 --compression_type=zstd --compression_zstd_max_train_bytes=0 \
--continuous_verification_interval=0 --db=/dev/shm/rocksdb/rocksdb_crashtest_whitebox \
--db_write_buffer_size=8388608 --delpercent=5 --delrangepercent=0 --destroy_db_initially=0 --enable_blob_files=0 \
--enable_compaction_filter=0 --enable_pipelined_write=1 --file_checksum_impl=big --flush_one_in=1000000 \
--format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=1000000 --get_property_one_in=1000000 \
--get_sorted_wal_files_one_in=0 --index_block_restart_interval=4 --index_type=2 --ingest_external_file_one_in=0 \
--iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True \
--log2_keys_per_lock=10 --long_running_snapshots=1 --mark_for_compaction_one_file_in=0 \
--max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000000 --max_key_len=3 \
--max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=16777216 --max_write_buffer_number=3 \
--max_write_buffer_size_to_maintain=8388608 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False \
--nooverwritepercent=0 --open_files=500000 --ops_per_thread=20000000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=1 --partition_pinning=0 --pause_background_one_in=1000000 \
--periodic_compaction_seconds=0 --prefixpercent=5 --progress_reports=0 --read_fault_one_in=0 --read_only=0 \
--readpercent=45 --recycle_log_file_num=0 --reopen=20 --secondary_catch_up_one_in=0 \
--snapshot_hold_ops=100000 --sst_file_manager_bytes_per_sec=104857600 \
--sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=False \
--target_file_size_base=2097152 --target_file_size_multiplier=2 --test_batches_snapshots=0 --test_cf_consistency=0 \
--top_level_index_pinning=0 --unpartitioned_pinning=1 --use_blob_db=0 --use_block_based_filter=0 \
--use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=0 --use_merge=0 \
--use_multiget=0 --use_ribbon_filter=0 --use_txn=0 --user_timestamp_size=8 --verify_checksum=1 \
--verify_checksum_one_in=1000000 --verify_db_one_in=100000 --write_buffer_size=4194304 \
--write_dbid_to_manifest=1 --writepercent=35
```

Reviewed By: pdillinger

Differential Revision: D27553054

Pulled By: riversand963

fbshipit-source-id: 60e391e4a2d8d98a9a3172ec5d6176b90ec3de98