Health Check for Object Storage
If you keep backups of your Mac computer in an object storage repository, you can schedule regular health checks to validate integrity of the backups in the repository.
Consider the following about health check for object storage:
- Veeam Agent verifies metadata of the whole backup, not just the latest restore point.
- Veeam Agent does not read data from data blocks in the storage; Veeam Agent only lists data blocks to make sure all blocks in the storage are available for rebuilding every restore point in the active backup chain. This mechanism reduces the number of requests to the storage, which makes health check for object storage cost-efficient.
If you want to run health checks for a backup that resides in an object storage repository, you must set a schedule according to which Veeam Agent will perform health checks. You can set the schedule in the Backup Job wizard or in command line interface to run health checks weekly or monthly on specific days.
When you configure a health check schedule, consider the following:
- Health check is run automatically during incremental backup job session on the days specified in the health check schedule. If the backup job runs several times on a specified day, health check is performed only with the first run of the backup job on that day.
Health check is not performed during the first full backup or subsequent active full backup job sessions.
- If Veeam Agent does not run any backup jobs on the day specified in the health check schedule, health check will be performed during the first backup job session following that day.
For example, you may have scheduled to run a health check every last day of a month, while the backup job is scheduled to run every day and to create an active full backup on Sundays. If the last day of a month falls on a Sunday, the health check will be performed on the following Monday with the first incremental backup job session on that day.
Veeam Agent performs a health check of a backup in the following way:
- During the backup job session after a new incremental backup file is created, Veeam Agent starts the health check of the whole backup. Veeam Agent checks if the metadata of the backup is consistent, and no metadata is missing. Veeam Agent also checks if all data blocks for every restore point are available on the storage. Veeam Agent does not read data from data blocks.
- If Veeam Agent does not find any corrupted data, the health check completes successfully. Otherwise, the health check completes with an error.
You can view the health check result in the session log. If during the health check, Veeam Agent finds corrupted data, it will also display information on where corrupt data has been detected — in backup metadata or data blocks, as well as list all restore points that share the corrupted data blocks.
Depending on the detected data inconsistency, Veeam Agent acts in one of the following ways:
- If the health check detects corrupted metadata, Veeam Agent will mark the backup chain as corrupted in the Veeam Agent configuration database; the backup job session will fail. During the next scheduled or manual backup job session, Veeam Agent will create a full backup and will start a new backup chain. The corrupted backup chain will become orphaned and will remain in the repository — you can keep or delete it.
- If the health check detects corrupted data blocks in the latest restore point of the active backup chain, Veeam Agent launches a health check retry.
During the health check retry, Veeam Agent will restart the backup job to create a new restore point. After that, Veeam Agent will transfer data blocks from the Veeam Agent computer to the backup repository. The transferred data blocks will include the blocks that were corrupted in the object storage repository and the blocks that changed since the start of the backup job session that triggered the health check. Veeam Agent will not perform another health check after the job retry is finished. The next health check will be run according to the defined schedule.
- If the health check detects corrupted data blocks in an inactive backup chain, Veeam Agent will not launch a health check retry. Veeam Agent will mark the backup and all related restore points as corrupted; the backup job session will end with a warning message.
If you try to restore data from a corrupted backup, Veeam Agent will display a warning message informing you that the restore operation may fail or the restored data may be corrupted.