perf(weave): split calls query when using a heavy condition #2779
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
|
Preview this PR with FeatureBee: https://beta.wandb.ai/?betaVersion=a6b55a606bcde7883323fb2fa7c2ed90ca572a2d |
| @@ -6,12 +6,27 @@ | |||
| from weave.trace_server.orm import ParamBuilder | |||
|
|
|||
|
|
|||
| def assert_sql(cq: CallsQuery, exp_queries, exp_params): | |||
| pb = ParamBuilder("pb") | |||
There was a problem hiding this comment.
every time I see this I think protobuf 🙃
| queries = cq.as_sql(pb) | ||
| params = pb.get_params() | ||
|
|
||
| for qr, qe in zip(queries, exp_queries): |
There was a problem hiding this comment.
queryRequest, queryExpected
queryReal?
tssweeney
approved these changes
Oct 29, 2024
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.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
WB-21659
This PR adds functionality to the CallsQuery builder to optionally generate two queries instead of one when conditions are right. If the query has a heavy filter or order by, we can't push down into the subquery, which causes a heinous join and bad performance. To remedy this we split the query into two parts, resolving the set of all possible call IDs, then adding that as a param to the param builder, then making a second query. This dance is managed by a helper function:
_construct_query_str_maybe_do_pre_querywhich always returns a valid sql query string. If we need to make a pre-query to get the IDs it happens in the helper, and returns the formatted next query.Testing
added test