Parity Rebuild next steps

[elphilo]

Hi all,

Over the weekend my aging RAID card seems to have died saying that two drives had to be disabled and multiple drives had read errors. Luckily this raid card was just passing through the drives to unRAID and I had an HBA card waiting in the wings.

 

Unfortunately the way the two cards present the drives to unRAID were different and I had to do a new config, which means I had to rebuild the parity drive. Luckily the majority of the drives seems to be fine as only one drive reported read errors, which were 128 total and one parity drive had to rewrite a few sectors, 5 total I believe.

 

With this being an aging system and me dragging my feet on doing what I’ve meant to do, getting new bigger hard drives, I pulled the trigger and ordered a bunch of drives to replace the aging drives I’ve been meaning to change.

 

Now my question is this. What do you think I should do next steps wise? My first thought was replace the drive that has read errors first, I’ve checked what’s on there and I’m not fussed if anything on that drive is “lost”. Then replace the drives unRAID said had to be disabled because of the failing RAID card. Then replace the parity. I also plan on setting up a NAS that’s been sitting around doing nothing in RAID 1 for a backup for my app/system/usb, but those drives don’t come in for a few more days.

 

I just wanted to see if anyone else had suggestions on what they would do or if I should do something else entirely.

 

Thanks!

[trurl]

Attach diagnostics to your NEXT post in this thread

[elphilo]

4 minutes ago, trurl said:

Attach diagnostics to your NEXT post in this thread

done

tower-diagnostics-20230410-0936.zip

[trurl]

You have a lot of disks, do any have SMART warnings on the Dashboard page?

[elphilo]

2 minutes ago, trurl said:

do any have SMART warnings on the Dashboard page?

Yeah, a few:

• Parity Drive 2- Reallocated sector count (45) & UDMA CRC error count (1)
• Disk 10 [The one with read errors] - Current pending sector (1)
• Cache Drive [I've known about this one for a while and have just been dragging my feet changing it] - Reallocated sector count (76)

 

My plan was Replace Disk 10, rebuild, Replace Parity, rebuild, back up Cache as hopefully by then I'll have the other drives I'm waiting for. Then slowly replace the ancient 2TB drives.

[trurl]

Since we don't have any diagnostics from when the problems occurred, no point in speculating about what happened. We will have to go from where you are now.

 

Except for that one pending sector on disk10, nothing seems critical unless reallocated are increasing. Other than that, SMART for all disks looks OK. Some have a lot of hours, and no SMART self-tests have been run. You might run extended self-tests on all of them while you wait on the new disks.

 

What size disks are you planning to replace then with? Parity is only 4TB, so you can't go larger wit data disks until you replace it first, and then parity2 is 6TB.

 

What is the unassigned 6TB sdd for?

[trurl]

Just now, trurl said:

run extended self-tests

You can do these tests on all disks at the same time.

[elphilo]

3 minutes ago, trurl said:

Since we don't have any diagnostics from when the problems occurred, no point in speculating about what happened. We will have to go from where you are now.

Yeah, I kind of freaked out when I saw two drives had been dropped and just decided to swap the old RAID card with the HBA card I've been meaning to for the past few months and hope for the best.

 

3 minutes ago, trurl said:

Except for that one pending sector on disk10, nothing seems critical unless reallocated are increasing. Other than that, SMART for all disks looks OK. Some have a lot of hours, and no SMART self-tests have been run. You might run extended self-tests on all of them while you wait on the new disks.

Yeah this server has been my main file server since 2011. I can start running those tests, but I now intend to change all the 2TBs out

 

3 minutes ago, trurl said:

What size disks are you planning to replace then with? Parity is only 4TB, so you can't go larger wit data disks until you replace it first, and then parity2 is 6TB.

I was going to replace all the 2TB with 4TB, the 6TB Parity with another 6TB parity. and then the 256GB cache with a 500GB cache drive.

 

3 minutes ago, trurl said:

What is the unassigned 6TB sdd for?

That should be another HDD, its for the torrent VM I have to download and seed torrents. I didn't want all those read/writes on the array and I don't really care for them being on the parity drive, hence their own drive.

 

So my initial plan of; Replace Disk 10, rebuild, Replace Parity, rebuild, back up Cache and replace as hopefully by then I'll have the other drives I'm waiting for. Then slowly replace the ancient 2TB drives; not a bad idea? 

[trurl]

Sounds good.

[trurl]

1 minute ago, trurl said:

Sounds good.

Results of self-tests might give more detailed priorities.

[elphilo]

Thanks for the help! I'm about to leave for work so I should probably start them before that.