Safest way to replace a working drive

[tony95]

My pool is an 8 drive Z2 and I want to replace a working drive. In my case, I had to use 1x 10TB drive in my pool along with 7x 8TB drives. So, now I need to do a replace on the 10TB drive and put an 8TB drive that I had been using in another computer into the array. This will degrade my pool until the re-silver finished. Is there any way to leverage the working drive to retain the 2 disk redundancy until the re-silver has completed? If I lost 2 of the remaining disks would the working drive I was replacing be of any use in recovering data or is it pretty much useless after the re-silvering begins? Is there an option other the replace that I don't know about? Thanks

 

[Etorix]

If you have a spare port, plug in the new drive without removing the old one. In GUI, go to Storage>Drives and note the device name of the drive you want to replace (adaN), then go to Storage>Pool>(gear)>Status>(drive)>(3-dot)>Replace. Resilver will take place with full redundancy.

 

[tony95]

Oh, so I just do the replace and don't remove the old drive until it finishes? Well that is simple. For some reason I was under the impression that when I started the replace that the old drive would be removed from the pool. Thanks

 

[danb35]

Yep, it really is that simple, as long as you have a spare SATA port. Just make sure the new disk is burned-in and tested first.

 

[Arwen]

@tony95 Their are rare cases when installing the replacement disk first, is less than ideal.

To sum it up, if you are replacing a "bad" disk, (not your case), and it has lots of bad blocks, the replacement process may waste too much time trying to read data from the disk to be replaced. Instead of simply using the redundancy of the pool, (RAID-Z2 in your case).

This replace in place can actually work on single drive pools. I've done it before to put in a larger disk.

In your case, all should be good.