fix: add buffer flushing to filesystem writes #1911
Merged
Conversation
This file contains 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
r3stl355
requested review from
wjones127,
fvaleye,
roeap and
ion-elgreco
as code owners
rtyler
approved these changes
Nov 29, 2023
Signed-off-by: Nikolay Ulmasov <[email protected]>
Signed-off-by: Nikolay Ulmasov <[email protected]>
ion-elgreco
force-pushed
the
issue-1770
branch
from
005f3cc
to
0192cf7
Compare
ion-elgreco
pushed a commit
to ion-elgreco/delta-rs
that referenced
this pull request
Dec 1, 2023
# Description
Current implementation of `ObjectOutputStream` does not invoke flush
when writing out files to Azure storage which seem to cause intermittent
issues when the `write_deltalake` hangs with no progress and no error.
I'm adding a periodic flush to the write process, based on the written
buffer size, which can be parameterized via `storage_options` parameter
(I could not find another way without changing the interface). I don't
know if this is an acceptable approach (also, it requires string values)
Setting the `"max_buffer_size": f"{100 * 1024}"` in `storage_options`
passed to `write_deltalake` helps me resolve the issue with writing a
dataset to Azure which was otherwise failing constantly.
Default max buffer size is set to 4MB which looks reasonable and used by
other implementations I've seen (e.g.
https://github.com/fsspec/filesystem_spec/blob/3c247f56d4a4b22fc9ffec9ad4882a76ee47237d/fsspec/spec.py#L1577)
# Related Issue(s)
Can help with resolving delta-io#1770
# Documentation
If the approach is accepted then I need to find the best way of adding
this to docs
---------
Signed-off-by: Nikolay Ulmasov <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Description
Current implementation of
ObjectOutputStreamdoes not invoke flush when writing out files to Azure storage which seem to cause intermittent issues when thewrite_deltalakehangs with no progress and no error.I'm adding a periodic flush to the write process, based on the written buffer size, which can be parameterized via
storage_optionsparameter (I could not find another way without changing the interface). I don't know if this is an acceptable approach (also, it requires string values)Setting the
"max_buffer_size": f"{100 * 1024}"instorage_optionspassed towrite_deltalakehelps me resolve the issue with writing a dataset to Azure which was otherwise failing constantly.Default max buffer size is set to 4MB which looks reasonable and used by other implementations I've seen (e.g. https://github.com/fsspec/filesystem_spec/blob/3c247f56d4a4b22fc9ffec9ad4882a76ee47237d/fsspec/spec.py#L1577)
Related Issue(s)
Can help with resolving #1770
Documentation
If the approach is accepted then I need to find the best way of adding this to docs