New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
upgrade to postgres 12 with jit off #4390
Conversation
Codecov Report
@@ Coverage Diff @@
## develop #4390 +/- ##
==========================================
Coverage 69.23% 69.23%
Complexity 3694 3694
==========================================
Files 267 267
Lines 15205 15205
Branches 1653 1653
==========================================
Hits 10527 10527
Misses 3909 3909
Partials 769 769
Flags with carried forward coverage won't be shown. Click here to find out more. Continue to review full report at Codecov.
|
No good |
nope, 12.0 is slow too |
Notes on factors to rule out via local testing:
|
Kudos, SonarCloud Quality Gate passed! |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Are you closing this PR in favor of the 13.x one?
@@ -42,7 +42,7 @@ services: | |||
# networks: | |||
# - elastic | |||
postgres_db: | |||
image: postgres:11.12 | |||
image: postgres:12.7 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Add command here as well?
No, the performance is slightly different and just in case of other issues we discover, I think I'll squash this one before the postgres 13 one so we can pick and choose whether to revert to 12 or 11 "next" time.
Undecided, I think I'll create a ticket to investigate when we get the performance testing back up. We're second guessing the query planner here and while that pays off in our tests, I'm not sure that is the case for "real". |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
FYI, I made this comment on https://ucsc-cgl.atlassian.net/browse/SEAB-3226, but I believe JIT is off by default in RDS.
companion for #4388
comparable performance to postgres 11 if we turn off jit
background at https://severalnines.com/database-blog/overview-just-time-compilation-jit-postgresql or hhttps://dba.stackexchange.com/questions/257759/recursive-cte-based-on-function-values-significantly-slower-on-postgres-12-than/258493#258493
guessing this might be a testing only issue since llvm should be available in the docker image, but we are inserting and dropping data rapidly which may be throwing off the query planner