SOBR rescan of capacity extent taking lots of time

[einhirn]

Hi everyone,

Case # 04463914

I'm in the process of moving Agent backups to our SOBR so they can make use of offloading/copying to object storage, too. Now, since I have to change the target repository on the Agent after moving it's backup data, I need to do this one by one, meaning I'll be rescanning the source repository and the SOBR quite a bit in the process.

But I ran into an issue with SOBR rescan taking 12 Hours for about 55TB of data. I'm using "copy" mode and don't keep backups in object storage any longer than on local disk - so both have about 850 vbk and about 5500 vib files. The performance tier gets rescanned in some minutes, but then it just sits there for hours, rescanning the on-premises, 10GbE-connected S3 storage. During this time I don't see any noteworthy load to explain the time it takes...

I repeated the rescan after the previous run finished successfully, but it even took a bit longer the second time around.

Is this to be expected? Perhaps some of you can share your rescan durations for SOBR with (on-premises) object storage?

Thanks and best regards!

[veremin]

Normally, rescan process should not take that longer, unless there are corrupted indexes on performance tier that need to be rebuild. However, we're constantly working on optimizing SOBR rescan performance, so, we will take a closer look at the logs provided to check whether there is issue we've missed. Thanks!