Optimize PagedInputStream::Skip #6699
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
✅ Deploy Preview for meta-velox canceled.
|
facebook-github-bot
added
the
CLA Signed
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D49501856 |
Yuhta
added a commit
to Yuhta/velox
that referenced
this pull request
Sep 25, 2023
Summary: Currently when we skip bytes in `PagedInputStream`, we do the decompression unconditionally and it is expensive. Some optimizations are added to address this: 1. Skip decompression of the whole block (frame in case of ZSTD) if 1. We can get the precise decompressed size, and 2. The decompressed size is no larger than the bytes need to skip 2. Accumulate contiguous skip calls to create larger skip region (delayed skipping) 3. Fix `ByteRleDecoder::skipBytes` to avoid reading data and breaking contiguous skips Differential Revision: D49501856
Yuhta
force-pushed
the
export-D49501856
branch
from
092254c to
d183dbc
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D49501856 |
Yuhta
added a commit
to Yuhta/velox
that referenced
this pull request
Sep 25, 2023
Summary: Currently when we skip bytes in `PagedInputStream`, we do the decompression unconditionally and it is expensive. Some optimizations are added to address this: 1. Skip decompression of the whole block (frame in case of ZSTD) if 1. We can get the precise decompressed size, and 2. The decompressed size is no larger than the bytes need to skip 2. Accumulate contiguous skip calls to create larger skip region (delayed skipping) 3. Fix `ByteRleDecoder::skipBytes` to avoid reading data and breaking contiguous skips Differential Revision: D49501856
Yuhta
force-pushed
the
export-D49501856
branch
from
d183dbc to
dfe16d4
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D49501856 |
Yuhta
added a commit
to Yuhta/velox
that referenced
this pull request
Sep 25, 2023
Summary: Currently when we skip bytes in `PagedInputStream`, we do the decompression unconditionally and it is expensive. Some optimizations are added to address this: 1. Skip decompression of the whole block (frame in case of ZSTD) if 1. We can get the precise decompressed size, and 2. The decompressed size is no larger than the bytes need to skip 2. Accumulate contiguous skip calls to create larger skip region (delayed skipping) 3. Fix `ByteRleDecoder::skipBytes` to avoid reading data and breaking contiguous skips Differential Revision: D49501856
Yuhta
force-pushed
the
export-D49501856
branch
from
dfe16d4 to
3f67f5b
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D49501856 |
Yuhta
added a commit
to Yuhta/velox
that referenced
this pull request
Sep 25, 2023
Summary: Currently when we skip bytes in `PagedInputStream`, we do the decompression unconditionally and it is expensive. Some optimizations are added to address this: 1. Skip decompression of the whole block (frame in case of ZSTD) if 1. We can get the precise decompressed size, and 2. The decompressed size is no larger than the bytes need to skip 2. Accumulate contiguous skip calls to create larger skip region (delayed skipping) 3. Fix `ByteRleDecoder::skipBytes` to avoid reading data and breaking contiguous skips Differential Revision: D49501856
Yuhta
force-pushed
the
export-D49501856
branch
from
3f67f5b to
408cab1
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D49501856 |
Summary: Currently when we skip bytes in `PagedInputStream`, we do the decompression unconditionally and it is expensive. Some optimizations are added to address this: 1. Skip decompression of the whole block (frame in case of ZSTD) if 1. We can get the precise decompressed size, and 2. The decompressed size is no larger than the bytes need to skip 2. Accumulate contiguous skip calls to create larger skip region (delayed skipping) Differential Revision: D49501856
Yuhta
added a commit
to Yuhta/velox
that referenced
this pull request
Sep 25, 2023
Summary: Currently when we skip bytes in `PagedInputStream`, we do the decompression unconditionally and it is expensive. Some optimizations are added to address this: 1. Skip decompression of the whole block (frame in case of ZSTD) if 1. We can get the precise decompressed size, and 2. The decompressed size is no larger than the bytes need to skip 2. Accumulate contiguous skip calls to create larger skip region (delayed skipping) Differential Revision: D49501856
Yuhta
force-pushed
the
export-D49501856
branch
from
408cab1 to
834c5c7
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D49501856 |
Yuhta
force-pushed
the
export-D49501856
branch
from
834c5c7 to
f0ea3ac
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D49501856 |
Contributor
|
This pull request has been merged in f6e9b76. |
Contributor
|
This pull request has been reverted by d08ab02. |
ericyuliu
pushed a commit
to ericyuliu/velox
that referenced
this pull request
Oct 12, 2023
Summary: Pull Request resolved: facebookincubator#6699 Currently when we skip bytes in `PagedInputStream`, we do the decompression unconditionally and it is expensive. Some optimizations are added to address this: 1. Skip decompression of the whole block (frame in case of ZSTD) if 1. We can get the precise decompressed size, and 2. The decompressed size is no larger than the bytes need to skip 2. Accumulate contiguous skip calls to create larger skip region (delayed skipping) Reviewed By: oerling Differential Revision: D49501856 fbshipit-source-id: 07241aaf71e83f0f491050a9be6075dd5500dd52
kletkavrubashku
added a commit
to kletkavrubashku/velox
that referenced
this pull request
Dec 23, 2025
Summary: ## History Some time ago similar PR was landed (facebookincubator#6699, D49501856) and caused SEV S369242 in Meta. That time `ZSTD_DCtx` context was created and reused on decompressor level. The optimization was reverted due to OOMs. ## Getting back to it again The optimization still makes sense. For example: - In Presto Adhoc we spend 0.07% of CPU cycles in `ZSTD_createDCtx_internal`: https://fburl.com/strobelight/2lt1wn4v - In Presto batch we spend 0.1% of CPU cycles in `ZSTD_createDCtx_internal`: https://fburl.com/strobelight/js4kn8za ## The fix Instead of creating `ZSTD_DCtx` per decompressor, we should create it per thread. Then we will be able to reuse the allocation and don't consume so much memory in FlatMaps. Differential Revision: D89716393
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Differential Revision: D49501856
Currently when we skip bytes in
PagedInputStream, we do the decompression unconditionally and it is expensive. Some optimizations are added to address this:ByteRleDecoder::skipBytesto avoid reading data and breaking contiguous skips