Is it better to migrate a RAID or to simply Rebuild?

I found out the answer to my own question via trial and error:

I have a TS-1685 with a five disk RAID5 and wanted to move to a six disk RAID 6. I installed the disk and started the migration with no problem. Well, there was one issue: The system reported that it would take about 530 hours to complete. To heck with that mess! I eliminated the original volume, after making sure that the data was backed up of course, and then started from the ground up creating a new RAID6 volume. Perfect. The system now reports it will be done in about 60 hours!

You also need to include the time taken to restore your data, and configure your shares and permissions again. :wink:

Couldn’t you just backup the system settings and then restore them after the rebuild? Or does that not include the shares and permissions data?

It does, but it also includes the RAID config.

If there was a way to selectively restore settings, that would be quite-useful.

Oh…Yeah. Then that would be a major pain to set everything up again!

Back to the OP: You could probably increase the raid re-build priority and it would take a lot less than 530 hours. Sure it would slow down the NAS a bit.

All good info for sure. Rebuild priority is at its highest and still am only getting about 100MBps while restoring the data at the save time. 50 hours and counting to RAID rebuild complete.

Raid rebuilding is very very slow as it basically has to copy and duplicate data sector by sector across all the drives. And now you have to have double parity which takes even longer.

530 hours is just 3 weeks. :smiley:

Hmm, just had a look at the system backup/restore process (I haven’t used it for many years) in QTS 5.2.5. It’s been updated and does now support selective restoration of setting groups.

I can’t see anything in there regarding RAID though. It’s not clear from the dialogs presented if RAID config is backed-up and restored.

Good-call @NA9D, this is an idea worth investigating further. :+1:

always start from scratch
to make sure all disks operate normally in stress pressure while raid build sync