Hello Saulo Paiva,
The issue has now been resolved. The container group that was previously returning InternalServerError
during deletion has been successfully removed by the Azure engineering team from the backend.
The root cause was a metadata inconsistency within the Azure Container Instances platform. A manual backend purge was necessary to remove it.
It was a backend fault within the Azure Container Instances platform. From what I can understand is the issue here is that the ACI resource is failing to delete due to an internal platform error, which Azure surfaces as an InternalServerError
. This could happen when there is a backend inconsistency or corruption in the container group metadata, or when orphaned dependencies like network profiles, DNS links, or virtual network attachments are incorrectly marked as still in use by the platform, even though the container itself is stopped and no locks are present. These stale references block the deletion operation and cannot always be resolved through the normal az container delete command or the Azure Portal.
From what I can see in the entire conversation so far is you've confirmed there are no resource locks on the container or resource group. You mentioned the container is not running and does not have any active services. You tried deleting through the Azure CLI and got the same internal server error. You waited over the weekend to rule out transient backend issues and you confirmed that you are the owner of the subscription and thus have full permissions.
At this point, my suggestion would be to try to bypass the CLI and portal layers and instead directly trigger the deletion through ARM REST API. You can use the az rest
command in Cloud Shell or your local terminal and you'll need to construct the full REST URL to the container group and call the DELETE
method against it.
Something like this-
az rest --method delete --url "https://management.azure.com/subscriptions/<your-subscription-id>/resourceGroups/<your-rg>/providers/Microsoft.ContainerInstance/containerGroups/<container-group-name>?api-version=2023-05-01"
This command calls the Azure ARM API directly and often resolves deletion issues by forcing a backend cleanup.
If this still fails, I recommend checking whether your container instance has an attached network profile by running
az container show --resource-group <your-rg> --name <container-group-name> --query "properties.networkProfile.id"
If a network profile is listed, it may be holding onto a virtual network reference that’s blocking the deletion. In that case, try deleting the network profile directly with
az network profile delete --ids <network-profile-id>
After removing it, retry the container group deletion. Let me know your findings. Based on that we can accordingly take next course of action. Thanks