Summary:
The patch adds a class called `BlobFileReader` that can be used to retrieve blobs
using the information available in blob references (e.g. blob file number, offset, and
size). This will come in handy when implementing blob support for `Get`, `MultiGet`,
and iterators, and also for compaction/garbage collection.
When a `BlobFileReader` object is created (using the factory method `Create`),
it first checks whether the specified file is potentially valid by comparing the file
size against the combined size of the blob file header and footer (files smaller than
the threshold are considered malformed). Then, it opens the file, and reads and verifies
the header and footer. The verification involves magic number/CRC checks
as well as checking for unexpected header/footer fields, e.g. incorrect column family ID
or TTL blob files.
Blobs can be retrieved using `GetBlob`. `GetBlob` validates the offset and compression
type passed by the caller (because of the presence of the header and footer, the
specified offset cannot be too close to the start/end of the file; also, the compression type
has to match the one in the blob file header), and retrieves and potentially verifies and
uncompresses the blob. In particular, when `ReadOptions::verify_checksums` is set,
`BlobFileReader` reads the blob record header as well (as opposed to just the blob itself)
and verifies the key/value size, the key itself, as well as the CRC of the blob record header
and the key/value pair.
In addition, the patch exposes the compression type from `BlobIndex` (both using an
accessor and via `DebugString`), and adds a blob file read latency histogram to
`InternalStats` that can be used with `BlobFileReader`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7461
Test Plan: `make check`
Reviewed By: riversand963
Differential Revision: D23999219
Pulled By: ltamasi
fbshipit-source-id: deb6b1160d251258b308d5156e2ec063c3e12e5e
Summary:
The patch adds support for writing blob files during flush by integrating
`BlobFileBuilder` with the flush logic, most importantly, `BuildTable` and
`CompactionIterator`. If `enable_blob_files` is set, large values are extracted
to blob files and replaced with references. The resulting blob files are then
logged to the MANIFEST as part of the flush job's `VersionEdit` and
added to the `Version`, similarly to table files. Errors related to writing
blob files fail the flush, and any blob files written by such jobs are immediately
deleted (again, similarly to how SST files are handled). In addition, the patch
extends the logging and statistics around flushes to account for the presence
of blob files (e.g. `InternalStats::CompactionStats::bytes_written`, which is
used for calculating write amplification, now considers the blob files as well).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7345
Test Plan: Tested using `make check` and `db_bench`.
Reviewed By: riversand963
Differential Revision: D23506369
Pulled By: ltamasi
fbshipit-source-id: 646885f22dfbe063f650d38a1fedc132f499a159
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:
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:
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:
After https://github.com/facebook/rocksdb/pull/7036, we still see extra DBTest that can timeout when running 10 or 20 in parallel. Expand skip-fsync mode in whole DBTest. Still preserve other tests from doing this mode to be conservative.
This commit reinstates https://github.com/facebook/rocksdb/issues/7049, whose un-revert was lost in an automatic
infrastructure mis-merge.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7274
Test Plan: Run all existing files.
Reviewed By: pdillinger
Differential Revision: D23177444
fbshipit-source-id: 1f61690b2ac6333c3b2c87176fef6b2cba086b33
Summary:
The patch makes sure that the functionality required for the new integrated
BlobDB implementation (most importantly, the classes related to reading and
writing blob files) is also built in LITE mode by removing the corresponding
`#ifndef`s.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7272
Test Plan: Ran `make check` in both regular and LITE mode.
Reviewed By: zhichao-cao
Differential Revision: D23173280
Pulled By: ltamasi
fbshipit-source-id: 1596bd1a76409a8a6d83d8f1dbfe08bfdea7ffe6
Summary:
BlobDB currently syncs each blob file periodically after writing a certain amount of
data (as specified by the configuration option `BlobDBOptions::bytes_per_sync`)
and all open blob files when the base DB's memtables are flushed. With the patch,
in addition to the above, blob files are also synced right before being closed, after
the footer has been written. This will be beneficial for the new integrated blob file
write path as well.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7160
Test Plan: `make check`
Reviewed By: riversand963
Differential Revision: D22672646
Pulled By: ltamasi
fbshipit-source-id: 62b34263543a7e74abcbb7adf011daa1e699998f
Summary:
Periodic syncing of blob files is performed by `WritableFileWriter`;
`bytes_per_sync_` and `next_sync_offset_` in `BlobLogWriter` are
actually unused (or more precisely, only used by methods that are
themselves unused). The patch removes all this dead code.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7125
Test Plan: `make check`
Reviewed By: riversand963
Differential Revision: D22531021
Pulled By: ltamasi
fbshipit-source-id: 6b293ad5a79d3e6bf15c5c68f7aedd7ce7a15f10
Summary:
After https://github.com/facebook/rocksdb/pull/7036, we still see extra DBTest that can timeout when running 10 or 20 in parallel. Expand skip-fsync mode in whole DBTest. Still preserve other tests from doing this mode to be conservative.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7049
Test Plan: Run all existing files.
Reviewed By: pdillinger
Differential Revision: D22301700
fbshipit-source-id: f9a9e3b3b26ce640665a47cb8bff33ba0c89b565
Summary:
The constant `kNoExpiration` is currently defined in an
internal/implementation header (`blob_log_format.h`); the patch moves it
to the public header `blob_db.h` so it is accessible to users.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7018
Test Plan: `make check`
Reviewed By: riversand963
Differential Revision: D22191354
Pulled By: ltamasi
fbshipit-source-id: 98c8012a83b999a3f1a30e955ce6bb71ba29dc5c
Summary:
The `FileMetaData` objects associated with table files already contain the
number of the oldest blob file referenced by the SST in question. This patch
adds the inverse mapping to `BlobFileMetaData`, namely the set of table file
numbers for which the oldest blob file link points to the given blob file (these
are referred to as *linked SSTs*). This mapping will be used by the GC logic.
Implementation-wise, the patch builds on the `BlobFileMetaDataDelta`
functionality introduced in https://github.com/facebook/rocksdb/pull/6835: newly linked/unlinked SSTs are
accumulated in `BlobFileMetaDataDelta`, and the changes to the linked SST set
are applied in one shot when the new `Version` is saved. The patch also reworks
the blob file related consistency checks in `VersionBuilder` so they validate the
consistency of the forward table file -> blob file links and the backward blob file ->
table file links for blob files that are part of the `Version`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6945
Test Plan: `make check`
Reviewed By: riversand963
Differential Revision: D21912228
Pulled By: ltamasi
fbshipit-source-id: c5bc7acf6e729a8fccbb12672dd5cd00f6f000f8
Summary:
The patch makes a couple of small cleanups to `SharedBlobFileMetaData` and `BlobFileMetaData`:
* It makes the constructors private and introduces factory methods to ensure these objects are always owned by `shared_ptr`s. Note that `SharedBlobFileMetaData` has an additional factory that takes a deleter object; we can utilize this to e.g. notify `VersionSet` when a blob file becomes obsolete (which is exactly when `SharedBlobFileMetaData` is destroyed).
* It disables move operations explicitly instead of relying on them being suppressed because of a user-declared destructor.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6749
Test Plan: `make check`
Reviewed By: riversand963
Differential Revision: D21206947
Pulled By: ltamasi
fbshipit-source-id: 9094c14cc335b3e226f883e5a0df4f87a5cdeb95
Summary:
The patch adds a couple of classes to represent metadata about
blob files: `SharedBlobFileMetaData` contains the information elements
that are immutable (once the blob file is closed), e.g. blob file number,
total number and size of blob files, checksum method/value, while
`BlobFileMetaData` contains attributes that can vary across versions like
the amount of garbage in the file. There is a single `SharedBlobFileMetaData`
for each blob file, which is jointly owned by the `BlobFileMetaData` objects
that point to it; `BlobFileMetaData` objects, in turn, are owned by `Version`s
and can also be shared if the (immutable _and_ mutable) state of the blob file
is the same in two versions.
In addition, the patch adds the blob file metadata to `VersionStorageInfo`, and extends
`VersionBuilder` so that it can apply blob file related `VersionEdit`s (i.e. those
containing `BlobFileAddition`s and/or `BlobFileGarbage`), and save blob file metadata
to a new `VersionStorageInfo`. Consistency checks are also extended to ensure
that table files point to blob files that are part of the `Version`, and that all blob files
that are part of any given `Version` have at least some _non_-garbage data in them.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6597
Test Plan: `make check`
Reviewed By: riversand963
Differential Revision: D20656803
Pulled By: ltamasi
fbshipit-source-id: f1f74d135045b3b42d0146f03ee576ef0a4bfd80
Levi Tamasi