Closed rlachic closed 2 months ago
@rlachic Thank you for reaching out to us. We've recently uncovered a bug which impacts revision creation. Having duplicate environment variables causes this bug. We have a fix and it's being deployed, but it will then reject your changes and make you choose one of the copies of your duplicate variable. As a workaround for now can you remove the duplicate environment variable.
Hi @simonjj , the problem with this workaround is the Container App stuck in a strange state, and we weren't able to even modify the variables. We tried several times, even activating older revisions, and didn't work. At least, the revision which doesn't work should be tag as unhealthy and give you the opportunity to re-balance to an older one if is possible.
Understood, out of curiosity this happens across portal, CLI and API? If redeployment is not an option I can see where this fix is in the deployment pipeline.
@simonjj I tried through Azure CLI and Portal, to create a new revision of the application without any environment variable, and the problem was the revision didn't be active because the stuck state of the Container App. So I was completely blocked.
This should be fixed now. No redeploy should be needed things should work again. Please reopen if you run into issues.
This issue is a: (mark with an x)
Issue description
Container App gets totally blocks when a new revision is wrong, and doesn't show any error, just let you with the previous version.
I have tried to stop the container and modify it, without success. If you stop the container, will not be able to run anymore.
I found the issue, and was related with a duplicated variable added through Azure CLI Or Azure API.
Anyway, this should report something, such as replica Unhealthy or something else, because doesn't show any kind of information and I have lost more than 1 day trying to figure out was going on.
This is the only error that I received through Azure CLI:
Failed to update revisions: Error activating revision something-app-backend-prod. Details Reason: Internal Server Error. Body: {"error":"Error updating revision","success":false}`
Steps to reproduce
Expected behavior [What you expected to happen.]
Don't get the container totally blocked. Show at least some kind of information such as replica unhealthy.
Actual behavior [What actually happened.]
Container is block, if the Container stop, doesn't run anymore.
Screenshots
If applicable, add screenshots to help explain your problem.
Additional context
Ex. Did this issue occur in the CLI or the Portal?