Use Parquet instead of Sequence files for entities #237
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
The original code was very fancy to avoid reading the input multiple times. This was a concern, for example with large CSVs. But now the input is always a Table, which is backed directly by a Parquet file in LynxKite storage. Parquet is column-oriented, so we can just read the columns one by one.
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Outdated suggestions cannot be applied.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.
This is a major change that we made within a private project. It had extensive code review there, and I'm just upstreaming it now. It's great. It simplifies the code, it fixes performance issues on GCS, it makes LynxKite easier to debug, and easier to integrate with external graph systems. It's faster too, by 10-20% on some benchmarks.
But it's an incompatible change for existing data. So this goes into LynxKite 5.0. But migration is simple: you just delete the
$KITE_DATA/partitioneddirectory. Things will be recomputed, so it may be slow, but at least it's easy.Copying one review comment and my response from the original PR:
I think when we originally wrote that code, it was possible for the DataFrame to be coming from anything. It could be a CSV, or a JDBC source. In that case it was important to do a single pass for performance / consistency. With the "new" (LynxKite 2.0) import procedure the table is always backed by a Parquet file.