Probably what I think... Just need verification...

[Binary Buddha]

So had a pool of five WD Red 8TB... But, I had to RMA one of the drives (the plastic data connection tab snapped off). So, now I'm running in a degraded Z2. But, I have a USB Seagate 8TB (SRD00F2) that is set as it's own pool that I was using for backups. Although, the backups are pretty out of date.

So, I'm thinking I could probably destroy the Seagate and add the drive to my main vdev while I wait for a replacement on the WD. The question is.... Would I be able to remove or replace the Seagate drive when the WD comes back and then retask the Seagate back to doing backups?

 

[nojohnny101]

Yes

 

[Binary Buddha]

Yes

To the ability to remove or replace?

 

[nojohnny101]

Sorry, yes and yes.

If I'm understanding you correctly, this situation would be no different then having a faulty drive and replacing it with a new one. When your replacement drive arrives, just follow the steps outline in the FreeNAS guide.

 

[BigDave]

So, I'm thinking I could probably destroy the Seagate and add the drive to my main vdev while I wait for a replacement on the WD.

The drive you mentioned is in an external enclosure, yes? If it's interface is USB, I would caution you strongly against doing this.
If the interface is eSATA that would be a better method of connection, if you have that type of port on your machine available.

 

[nojohnny101]

Sorry I just reread your original post. I thought you were going to remove the 8tb from the enclosure and then attach it internally. If so all I said applies.

If you want to use the 8tb and add it to the vdev while keeping it in its enclosure then @BigDave is correct, highly not recommended.

 

[BigDave]

I just re-read the OP and had another word of caution. If the 8TB external has a backed up copy of your data, don't touch
it at this point. If you reformat THAT drive and something goes wrong, where's your backup???

 

[Binary Buddha]

I just re-read the OP and had another word of caution. If the 8TB external has a backed up copy of your data, don't touch
it at this point. If you reformat THAT drive and something goes wrong, where's your backup???

Yeah, but the backup that's on it is out of date. Doesn't have the new datastore reconfigs that I did. (After I learned about the beauty of datastores.)

 

[joeschmuck]

Id' just shut down the server and wait for the new drive to arrive if that is at all possible.

 

[BigDave]

Yeah, but the backup that's on it is out of date. Doesn't have the new datastore reconfigs that I did. (After I learned about the beauty of datastores.)

That's unfortunate, but it is what it is. An older backup is better than the possibility of no backup at all.

Id' just shut down the server and wait for the new drive to arrive if that is at all possible.

+1 for joeschmuck's suggestion!!!

 

[danb35]

What's keeping you from making a more recent backup?

 

[Stux]

Essentially, you're just back to having raidz1.

Refresh your backup. If you still are sweating bullets, then buy another HD.

(I've done exactly this when RMAing an array member)

Then when you get the RMA back it can be used as part of your backup store ;)

 

[Binary Buddha]

What's keeping you from making a more recent backup?

um... laziness...

Relocated the server so now I gotta hookup power to the drive and all that jazz.

 

[Mirfster]

um... laziness...

Relocated the server so now I gotta hookup power to the drive and all that jazz.

Wouldn't/Shouldn't that already be part of your regular backup routine?

 

[rs225]

You don't want to use a USB drive in an array.

An 8TB Seagate in a USB case has a strong chance of being an Archive/SMR drive, which also has no business being in a raidz.

Freshen your backup.

Don't sweat the missing drive; raidz1 still has 99% the failure-fighting power of raidz2. User error, however, is fatal for any pool.

 

[Binary Buddha]

Wouldn't/Shouldn't that already be part of your regular backup routine?

Was... Until I relocated...


Sent from my iPhone using Tapatalk

 

[Binary Buddha]

User error, however, is fatal for any pool.

User error? Examples?


Sent from my iPhone using Tapatalk