Python: Fix Avro scan performance in PyArrow #6283
Merged
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
rdblue
commented
Nov 27, 2022
| read_bytes = self._input_stream.read(n) | ||
| if len(read_bytes) != n: | ||
| read_len = len(read_bytes) | ||
| if read_len <= 0: |
Contributor
Author
There was a problem hiding this comment.
In Python, EOF is signaled by returning 0 bytes.
rdblue
commented
Nov 27, 2022
| self.reader = visit(self.schema, ConstructReader()) | ||
| else: | ||
| self.reader = resolve(self.schema, self.read_schema) | ||
| self.reader = cast(StructReader, resolve(self.schema, self.read_schema)) |
Contributor
Author
There was a problem hiding this comment.
This fixes a type warning.
Fokko
approved these changes
Nov 28, 2022
Contributor
Fokko
left a comment
Fokko
left a comment
There was a problem hiding this comment.
This is awesome, thanks for figuring this out! 🚀
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Outdated suggestions cannot be applied.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.
This improves Avro scan performance when using PyArrowFileIO by about 10x by ensuring that scans are buffered and updating EOF handling. I was testing scan planning and found that S3 planning took about 130 seconds. Buffering the input stream and avoiding the
len(input_file)call gets the planning time for the same query to 13 seconds.The main problem was that the stream provided by PyArrow was not buffered, so nearly every read operation was causing another request to S3.
A second issue was that the call to get the file length was expensive, so this also updates the decoder to throw
EOFExceptionrather than checking file length. This improved the performance by about 1 second in my test.