azure/cli@v2 started giving error for az containerapp revision copy

[mehdi-eaton]

Describe the bug

Hi Team,

Since morning, we have observer our pipeline failing.
I am using github action with:

• name: Azure CLI script
  uses: azure/cli@v2
  env:
  containerAppName: 'xxx'
  resourceGroup: abc
  with:
  azcliversion: latest
  inlineScript: |
  az containerapp revision copy -n ${{ env.containerAppName }}
  -g ${{ env.resourceGroup }} --revision-suffix ${{ github.run_id }}-${{ github.run_attempt }}

This was working till yesterday.

I have also verified same command If I run from my local machine az cli its working fine.

Related command

• name: Azure CLI script
  uses: azure/cli@v2
  env:
  containerAppName: 'xxx'
  resourceGroup: abc
  with:
  azcliversion: latest
  inlineScript: |
  az containerapp revision copy -n ${{ env.containerAppName }}
  -g ${{ env.resourceGroup }} --revision-suffix ${{ github.run_id }}-${{ github.run_attempt }}

Errors

Issue script & Debug output

Expected behavior

Container app new revision should have created.

Environment Summary

Running on Github with action azure/cli@v2 with azcliversion: latest

Additional context

No response

[yonzhan]

Thank you for opening this issue, we will look into it.

[mehdi-eaton]

This is fixed after added :

some how using latest tag is pulling 2.69.0

[scott-wilkos]

We started receiving this same error, with the "containerapp ........ not found" starting the first time 5/15/2025 around 05:30 EST. This was after no changes in a year for the YML file or changes to the Azure environment. We even provision ahead of this step, so even if it was deleted it would have been re-created.

Currently has us unable to deploy and I'd be glad to provide any supporting documentation to assist with this.

[AHarman]

We've found this happening specifically in Azure China in version 2.72.0. Using 2.71.0 works in China, using 2.72.0 works in regular Azure.

I saw some discussion somewhere about a change to the underlying API method, but I'm struggling to find that now. here: hashicorp/terraform-provider-azurerm#29317. Maybe this has something to do with it?

Around this time there have been updates to the version of the Azure API being used. Is it possible that the sovereign clouds are returning 404 instead of 400? Normally they return an error like "unexpected status 400 (400 Bad Request) with error: NoRegisteredProviderFound" . Were the endpoints changed or something?

[Greedygre]

The error "The containerapp '{}' does not exist" only happened when getting the containerapp failed.
@mehdi-eaton
Can you ensure the containerapp exists or use --debug to validate the result of the request to Get the containerapp?

I created PR: #31689 to avoid print containerapp does exists when response is not 404

Thanks