DM-43205: Calibs fail to locally certify in Prompt Processing #135
Conversation
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
For some reason, duplicate calibs are getting preloaded. This should help find them.
kfindeisen
force-pushed
the
tickets/DM-43205
branch
from
acc198a to
eaa5d4a
Compare
hsinfang
reviewed
Mar 11, 2024
kfindeisen
force-pushed
the
tickets/DM-43205
branch
from
eaa5d4a to
7b7c5be
Compare
The option prevents crashes when calibs are fed out of chronological order to our validity workaround code. It should not be set for any production run.
The S3 code always tries to flush even though flushing is not supported for small files; the file gets written correctly, but the user can't stop the attempt.
kfindeisen
force-pushed
the
tickets/DM-43205
branch
from
2c4f61a to
d25fcea
Compare
hsinfang
approved these changes
Mar 13, 2024
|
|
||
| # VALIDITY-HACK: local-only calib collections break if calibs are not requested | ||
| # in chronological order. Turn off for large development runs. | ||
| cache_calibs = bool(int(os.environ.get("DEBUG_CACHE_CALIBS", '1'))) |
There was a problem hiding this comment.
All other parameters we load from environment are in activator.py. Would it be easier to manage if this goes there too?
There was a problem hiding this comment.
That's... not actually true. There are quite a few calls to os.environ here, I was just lazy and buried the others in the code. 😬
If the variables got moved to activator.py, how would we make them accessible to this module's code? There's too many to be good parameters to the MiddlewareInterface constructor, and in any case a lot of them (including cache_calibs) are Middleware implementation details that activator shouldn't know about.
This was referenced Jan 23, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
This PR modifies the calib validity hack introduced on #129 to optionally purge the calib cache, preventing conflicts if two calibs are loaded on the same day. While calibs do not change mid-day in production, our integration tests (especially HSC RC2) mix a wide range of "real" observing dates in a single run, and the previous hack would break them.