Fix plan cost runner regression script #11129
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
Tmonster
force-pushed
the
fix_plan_cost_runner_regression
branch
from
936fa1d to
2b45c30
Compare
|
Thanks! |
github-actions bot
pushed a commit
to duckdb/duckdb-r
that referenced
this pull request
Mar 21, 2024
Merge pull request duckdb/duckdb#11276 from pdet/error_handler_csv Merge pull request duckdb/duckdb#11129 from Tmonster/fix_plan_cost_runner_regression
krlmlr
added a commit
to duckdb/duckdb-r
that referenced
this pull request
Mar 23, 2024
Merge pull request duckdb/duckdb#11276 from pdet/error_handler_csv Merge pull request duckdb/duckdb#11129 from Tmonster/fix_plan_cost_runner_regression
github-actions bot
pushed a commit
to duckdb/duckdb-r
that referenced
this pull request
Mar 28, 2024
Merge pull request duckdb/duckdb#11276 from pdet/error_handler_csv Merge pull request duckdb/duckdb#11129 from Tmonster/fix_plan_cost_runner_regression
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Outdated suggestions cannot be applied.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.
Thijs had some comments on the old PR https://github.com/duckdb/duckdb/pull/10585/files. I've been messing around in the Join Order Optimizer and realized that I old got my equalities incorrect and the regression / improvement reporting was flipped 🙈 .
This PR also removes comparing build side and probe side for plans and reporting a regression if any of these values increase. If the total plan cost changes, these values more than likely change as well, and I want to avoid being eager about reporting regressions especially if the execution time has improved. The plan now is only to report changes if the total plan cost has increased and the build side has increased. When this is not the case a regression is reported if there is a significant (3%) difference in execution time. It's possible that build side and probe side cardinalities have changed, but execution time has improved. For example, there may be more intermediate tuples in the join tree, but these tuples might all be in the probe side. This won't add much overhead since the tuples will be in flight most of the time. If the tuples were on the build side, that would add a overhead since they will need to be included in the hash table.
I ran this new version plan_cost_runner.py between v0.9.2 and the current main and there are no regressions for tpch and imdb.