Hi all,
I'm currently trying to duplicate a SQL Server EE database following the instructions here: https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/SQLServer.Procedural.Importing.html
The original db looks like this:
1> use OLDDB
2> exec sp_spaceused
3> go
Changed database context to 'OLDDB'.
database_name database_size unallocated space
-------------------------------------------------------------------------------------------------------------------------------- ------------------ ------------------
OLDDB 16796.31 MB 5271.73 MB
reserved data index_size unused
------------------ ------------------ ------------------ ------------------
1870552 KB 1685304 KB 165616 KB 19632 KB
I also have:
D:\rdsdbdata\DATA\OLDDB.mdf = 6GB
D:\rdsdbdata\DATA\OLDDB_log.ldf = 9GB
The backup that lands in the S3 bucket is ~ 1.8GB and completes without error, but the restore task doesn't and ends in the state:
5 RESTORE_DB_NORECOVERY NEWDB 0 1 ERROR
[2023-09-27 15:44:22.387] Aborted the task because of a task failure or a concurrent RESTORE_DB_NORECOVERY request.
[2023-09-27 15:44:22.453] Task has been aborted
[2023-09-27 15:44:22.457] Maximum supported database size on SQL Server Express edition is: 2023-09-27 15:44:22.457 2023-09-27 15:43:36.700 arn:aws:s3:::mybucket
The RDS instance (source and target) is a db3.t.large (2CPU/8GB) with 200GB, RDS version 15.00.4198.2.v1, my backup command is:
exec msdb.dbo.rds_backup_database @source_db_name='OLDDB',
@s3_arn_to_backup_to='arn:aws:s3:::mybucket/olddb.bak',
@overwrite_s3_backup_file=1, @type='FULL', @number_of_files=1;
and restore:
exec msdb.dbo.rds_restore_database @restore_db_name='NEWDB',
@s3_arn_to_restore_from='arn:aws:s3:::mybucket/olddb.bak',
@with_norecovery=1, @type='FULL';
I went through the troubleshooting steps in https://repost.aws/knowledge-center/rds-sql-server-fix-native-backup-restore, but I can't find any obvious mistakes in my configuration, policy and policy-assignments all look ok.
Am I running into some limitation (due to EE) here or am I doing something wrong?
Thanks,
Marc
Edit: FWIW, I can backup and restore a much smaller database on the same instance using the exact same process without a problem.
Adrian's recommendation would probably be the best way to resolve this, unfortunately the db owner choose to use a workaround to copy their db and leave the original as it is. I don't know if the inability to restore backups is the only consequence the exceeded file size limit has. Looks like we might find out eventually.