-
Notifications
You must be signed in to change notification settings - Fork 1.1k
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
Clear scalers cache correctly both in Operator and Metrics Server #2564
Conversation
Signed-off-by: Zbynek Roubalik <[email protected]>
Signed-off-by: Zbynek Roubalik <[email protected]>
/run-e2e |
@zroubalik I've just tested the change I am working on (#2495) against this PR. This PR resolves the issue I described in Slack (the metrics server -- but not the operator -- failing to reload secrets in response to auth failures). I now see one auth failure in the logs at credential expiration time, followed by success once the creds have been reloaded. |
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.
LGTM!
Awesome bug fix for the release!! Congrats! 🥇🥇🥇
…dacore#2564) Signed-off-by: Zbynek Roubalik <[email protected]> Signed-off-by: Mark Rzasa <[email protected]>
…dacore#2564) Signed-off-by: Zbynek Roubalik <[email protected]> Signed-off-by: Mark Rzasa <[email protected]>
author Mark Rzasa <[email protected]> 1642970493 -0500 committer Mark Rzasa <[email protected]> 1643317483 -0500 app insights scaler Signed-off-by: Mark Rzasa <[email protected]> Add PredictKube scaler (kedacore#2418) Signed-off-by: Daniel Yavorovych <[email protected]> Signed-off-by: alex60217101990 <[email protected]> bump deps (kedacore#2563) Signed-off-by: Zbynek Roubalik <[email protected]> Clear scalers cache correctly both in Operator and Metrics Server (kedacore#2564) Signed-off-by: Zbynek Roubalik <[email protected]> test: Provide e2e secrets for Azure Application Insights scaler (kedacore#2568) Signed-off-by: Tom Kerkhove <[email protected]>
author Mark Rzasa <[email protected]> 1642970493 -0500 committer Mark Rzasa <[email protected]> 1643317483 -0500 app insights scaler Signed-off-by: Mark Rzasa <[email protected]> Add PredictKube scaler (kedacore#2418) Signed-off-by: Daniel Yavorovych <[email protected]> Signed-off-by: alex60217101990 <[email protected]> bump deps (kedacore#2563) Signed-off-by: Zbynek Roubalik <[email protected]> Clear scalers cache correctly both in Operator and Metrics Server (kedacore#2564) Signed-off-by: Zbynek Roubalik <[email protected]> test: Provide e2e secrets for Azure Application Insights scaler (kedacore#2568) Signed-off-by: Tom Kerkhove <[email protected]> Signed-off-by: Mark Rzasa <[email protected]>
Signed-off-by: Zbynek Roubalik [email protected]
This fix should correctly clear/invalidate cache in both KEDA Operator and Metrics Server, when ScaledObject is deleted from cluster, when there is a problem in resolving secrets/credentials or when there is a general problem in scraping metrics from a scaler.
The previous fix #2408 worked only partially - in KEDA Operator, because it didn't use
ClearScalersCache()
function, but tried to delete the map entry directly. But the main problem was, that theClearScalersCache()
which is being used in Metrics Server, operated on a wrong key, ie.name.namespace
instead ofkind.namespace.name
generated bywithTriggers.GenerateIdenitifier()
. That resulted in cache not being invalidated in Metrics Server.Checklist
Fixes #2407