Re-activating a previous revision in single revision mode

[bowlerma]

Please provide us with the following information:

This issue is a: (mark with an x)

• [x] bug report -> please search issues before submitting
• [ ] documentation issue or request
• [ ] regression (a behavior that used to work and stopped in a new release)

Issue description

In the Azure portal re-activating a previous revision in single revision mode fails if the container registry is in a different subscription.

Steps to reproduce

1. Deploy a revision using the "Docker Hub or other registries" option to select an Azure Container Registry in another subscription. (workaround for issue where you can't select Container Registries in another subscription using the "Azure Container Registry' option.)
2. Repeat step 1 to deploy a new revision.
3. From "inactive Revisions" select the previous revision and click "Activate"
4. Click "Save" to deploy the previous revision.

Expected behavior [What you expected to happen.] The container app should revert to the previous revision with no downtime.

Actual behavior [What actually happened.] The following error is reported from the notifications:

Failed to update revisions: Error activating revision <redacted>--dgs0s4f. Details Reason: Bad Request. Body: {"error":"Error updating revision <redacted>--dgs0s4f. You may be facing this error because this revision references a registry which is no longer defined. Please check registry definition and try again.","success":false}

The activity log shows the following in the JSON:

    "status": {
        "value": "Failed",
        "localizedValue": "Failed"
    },
    "subStatus": {
        "value": "Conflict",
        "localizedValue": "Conflict (HTTP Status Code: 409)"
    },

The current revision then goes into a failed state whilst the container restarts and probes indicate that the container is ready again which results in the container app being unavailable.

Additional context

Issue occurs in the portal. Not tried the CLI.

[howang-ms]

We have identified this is a bug, and we are working on the fix.

[dsczltch]

@howang-ms We also have this issue and are waiting for a bugfix. Could you please communicate your ETA?