Conversation
brainrake
approved these changes
Sep 30, 2026
Collaborator
|
Buys us until next tz change to get rolling updates going. |
This comment has been minimized.
This comment has been minimized.
Collaborator
Author
|
Result from quick build/test: |
mmlb
force-pushed
the
mannymendez/mpg-1289-update-postgres-tzdata-for-albertas-permanent-utc-6-change
branch
from
October 1, 2026 13:27
fa273ee to
29196e6
Compare
|
Blacksmith runners detected OOM events on the following jobs:
|
mmlb
force-pushed
the
mannymendez/mpg-1289-update-postgres-tzdata-for-albertas-permanent-utc-6-change
branch
from
October 1, 2026 17:15
29196e6 to
14c7bc6
Compare
Collaborator
Author
|
GHA wasn't happy with the nix-eval step and I couldn't quite debug it from logs available. I had some not-quite-working code to split the run nix-eval-jobs, process nix-eval-jobs output in github-matrix into 2 parts that I cleaned up/finished and now things are ok. Will keep it in. |
Better to save/load the expensive step (nix-eval-jobs) instead of some intermediate step after further processing. Can't save via stdout because the GHA notice/warning/... functions have send to stdout themselves (because GHA doesn't read from stderr) which makes for invalid json.
The eval/gen-matrix step is not currently working without much info to go on. I suspect that since I'm doing a change to the base postgres derivation and causing a mass rebuild it generates a huge json that likely exceeds GHA's max payload for ::notice:: messages. All I can see is that the notice message printed to stdout is truncated with plenty of more chars left to print. So lets rework the script so the important part comes first, setting the outputs then we can inform any watchers/readers of what needs building. I switched to just dumping to stdout because that large of a notice isn't really necessary in the GH UI, log file is fine.
This all happens on an ephemeral runner, no need to stop nix daemon or much of anything else being done. There's no pipes so no need for pipefail, no bash vars so no need for -u, no subshells so no need for -E and .#blah is not a shell comment so no need to escape.
Not sure whats going on in CI, logs are being truncated and the matrix isn't being populated. I thought the previous commit setting outputs first would be helpful but that doesn't seem to be the case. I suspect nix-eval-jobs isn't allowed to finish or something so lets break the execution up.
Lets let PG use its own copy of tzdata instead of nixpkgs since we're likely to be faster at updating PG over nixpkgs going forward. Morocco is already on permanent UTC+00 and Alberta will be permanent UTC-06 in November, PG's tzdata was updated[1] but our nixpkgs' tzdata has not and will not in a timely fashion. [1]: https://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=669438aed
mmlb
force-pushed
the
mannymendez/mpg-1289-update-postgres-tzdata-for-albertas-permanent-utc-6-change
branch
from
October 1, 2026 17:56
14c7bc6 to
8d07184
Compare
mmlb
deleted the
mannymendez/mpg-1289-update-postgres-tzdata-for-albertas-permanent-utc-6-change
branch
October 2, 2026 13:27
This file contains hidden or 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
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.
What kind of change does this PR introduce?
Feature
What is the current behavior?
PG is built against nixpkgs' tzdata, which is stale and will remain so for a bit still.
What is the new behavior?
PG is built to use its own embedded tzdata which is up to date.