- Mais recentes
- Mais votos
- Mais comentários
I feel just that in case of Aurora Cluster restored from restore point only restore the Aurora cluster. You need to manually attach the instances with the cluster. Second would be if you can see the KMS used for encryption of source Aurora has permission's granted to the account you are trying to perform the restore.
You can look at CloudTrail event history around the time when you run the restore to see if any CreateDBInstance call is failing/giving an error. The error code would tell the missing permission.
Please also look at any other API calls around that timestamp to note any additional API calls which gave an error.
Hello Shivam, I did and I got no errors, I see the following actions: StartRestoreJob, RestoreDBClusterFromSnapshot, CreateGrant, RestoreStarted
I could not find any error in CloudTrail, the Bakcup Restore Job shows as Completed, the RDS Cluster is there with status of Available but with 0 instances.
Conteúdo relevante
- AWS OFICIALAtualizada há 10 meses
- AWS OFICIALAtualizada há um ano
- AWS OFICIALAtualizada há um ano
I will make some tests and update. I think the restore from AWS Backup should create at least one instance, doesn't make sense to restore to cluster with 0 instances, in that case as I do now, I prefer to restore from Snapshot from the RDS Consosole.
Update: exactly as @SSaxena said, the Aurora Cluster is created without Instance, you havce to manually add a new Instance, I think I will stick to the restore from the RDS Console for now