Open Rick-healy opened 1 month ago
@Rick-healy Hello, thanks for raising this. I was able to recreate the problem some of the times. I believe this may be an issue with service processing requests for 2 backup resources in parallel. I will confirm with the service team on this, in the mean time, I believe adding the first back up resource as a dependency for the second back up resource in the depends_on
property should solve the problem.
Let me know any concerns, further problems.
Thank you @hqhqhqhqhqhqhqhqhqhqhq for responding. Yes it does sound like some parallel issue. My actual code has the instances being created in a for-each loop from an object map, so I don't think I can add dependencies in the loop for each iteration, but I can investigate other non-parallel approaches for this. Or add the instances one at a time in the TF and re-run, which isn't a massive overhead, as they aren't going to be too dynamic!
We are seeing the same error attempting to create a backup instance in a backup vault that has an existing soft deleted instance
We are seeing the same error attempting to create a backup instance in a backup vault that has an existing soft deleted instance
-Ran into the same issue when redeploying a backup vault. Were you able to remove the soft deleted instance or find a workaround?
Same issue when creating a backup vault from scratch
Error: creating/updating DataProtection BackupInstance ("Backup Instance (Subscription: \"xxxxxxxxxxxxxx\"\nResource Group Name: \"xxxxxx\"\nBackup Vault Name: \"xxxxxx\"\nBackup Instance Name: \"xxxx\")"): performing CreateOrUpdate: unexpected status 406 (406 Not Acceptable) received with no body
Same issue here
Is there an existing issue for this?
Community Note
Terraform Version
4.241.26
AzureRM Provider Version
3.35.0
Affected Resource(s)/Data Source(s)
azurerm_data_protection_backup_instance_kubernetes_cluster
Terraform Configuration Files
Debug Output/Panic Output
Expected Behaviour
I'd expect 2 each of the backup instances to be created.
Actual Behaviour
When attempting to create more than one Kubernetes backup instance in Terraform (whether they are separate resource blocks or a for-each loop, the 'creating' process initiates, however as soon as the first instance reports completed, the process stops with http error code 406, and any remaining instances still being created fail.
If you re-run the terraform unchanged, then again, one instance will be created, but if there is more than one instance still to be created then the same error occurs.
You can repeat this until all are created.
The same error occurs in the destroy process.
Steps to Reproduce
Run the TF with the multiple backup instances specified. It will start to create all the instances, but as soon as the first one reports 'complete' the others will all fail with 406 error.
Important Factoids
No response
References
No response