Speed up CollateProducts by avoiding redistribute when possible.
#191
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
ljgray
changed the title
CollateProducts by avoiding redistrubute when possible.CollateProducts by avoiding redistribute when possible.
jrs65
requested changes
Sep 6, 2022
ljgray
force-pushed
the
ljg/collate-products-va
branch
from
02d803c
to
2df2016
Compare
|
@jrs65 Improved the test for frequency mapping, this should be good to merge now |
jrs65
approved these changes
Sep 7, 2022
In many cases, the frequecy mapping does not reorder the frequency axis at all, and there is no need to redistribute away from the frequency axis. The redistribute call is expensive, so we avoid it when possible.
jrs65
force-pushed
the
ljg/collate-products-va
branch
from
2df2016
to
1ff6bf4
Compare
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Outdated suggestions cannot be applied.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.
The call to redistribute is only needed if the frequency axis has to be reordered. In many cases this doesn't have to happen, so we can do the task while distributed across frequency, avoiding an expensive
redistributecall. This checks to see if the frequency mapping is sorted, and if it is it does not redistribute. Without redistributing, the task takes 70-80s total in the daily pipeline, down from around 2500s when redistributing every time.