[SPARK-14259] [SQL] Merging small files together based on the cost of opening #12095
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
|
ping @nongli |
|
Test build #54661 has finished for PR 12095 at commit
|
|
cc @marmbrus |
|
Lgtm, though I still think pipelined reading is the right thing to do long term. |
|
Pipelined reading is a good idea to overlap opening and processing, will be done in another PR. After that, this PR is still useful, thinking of lots of tiny files (100ms to open, but 1ms to process), opening these files will be majority of the time, we should not add thousands of tiny files into one partition. |
|
Merging this into master, thanks! |
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Outdated suggestions cannot be applied.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.
What changes were proposed in this pull request?
This PR basically re-do the things in #12068 but with a different model, which should work better in case of small files with different sizes.
How was this patch tested?
Updated existing tests.
Ran a query on thousands of partitioned small files locally, with all default settings (the cost to open a file should be over estimated), the durations of tasks become smaller and smaller, which is good (the last few tasks will be shortest).