I've configured one backup to tape job for all our Backup jobs and repositories - VMware and Veeam Agents. When this runs, I often see a troughput graph like this:
this shows that the job could be faster if it would better use the multiple drives available. I reach this conclusion because we have some backup chains in the multi terabyte range, and some in the few hundred megabyte range. The graph shows all drives being used at first, but in the end there's one big chain that is using only one drive. If we start processing with these big chains and continue with the smaller chains, I guess more drives would be kept from idling longer.
So please consider this request:
When a tape job includes more than one backup chain, sort those backup chains by size descending before processing them. That way all drives will be kept busy and the throughput will stay high for the maximum amount of time.
Thanks!
Kind regards,
einhirn