Failed hardware, moved to new and now have multiple missing drives

[Steiner49er]

Hi everyone,

 

Sadly I had a server that just decided to up and stop powering on at all, out of warranty of course and I don't have the funds to buy components to swap out hardware to see what failed (most likely motherboard I think).

 

I've swapped from the (dead) Dell R730XD server to an HP DL380 Gen8 server, but Unraid isn't able to automatically recognize the drives, or at least most of them.  I can work out what some of them are, but when I do select that it says "Wrong" now instead of missing.  For the ones that I can't identify, because I have several WD 8TB drives, what can I do?  I had two WD 8TB parity drives, and data drives consisting of three WD 8TB drives, one Seagate 4TB drive, and several other 1.2-1.8 SAS drives, some of those 1.2's I couldn't move into the HP server as I was missing enough drive trays to move them all.

 

I did try to search the forums but didn't come up with something similar, likely my fault in search terms I think!

 

I do have the HP P420i array controller in HBA mode thanks to another post on the forums, so I should be set there, but I think due to the controller differences it's getting more(?) information from the drives and that doesn't match what it had previously. 

[ChatNoir]

Please attach your diagnostics to your next post.

 

Do you have an idea of your drives allocation ? Maybe a recent daily array health report ?

[Steiner49er]

22 minutes ago, ChatNoir said:

Please attach your diagnostics to your next post.

 

Do you have an idea of your drives allocation ? Maybe a recent daily array health report ?

Hmm... not that I know of on the daily array health report, is there somewhere I can look that it might be stored?  I do have the server in My Servers if there's anything that may be there?

 

Diagnostics attached.

tower-diagnostics-20220911-1617.zip

[trurl]

Post a screenshot of Main - Array Devices and Main - Pool Devices

[Steiner49er]

30 minutes ago, trurl said:

Post a screenshot of Main - Array Devices and Main - Pool Devices

I'm going to have to reboot the server I guess, this happens once in a while, including on the old hardware as well, the drives are there but everything is blank.

[itimpi]

54 minutes ago, Steiner49er said:

Hmm... not that I know of on the daily array health report, is there somewhere I can look that it might be stored?  I do have the server in My Servers if there's anything that may be there?

In the worst case the approach detailed here in the the online documentation accessible via the ‘Manual’ link at the bottom of the GUI or the DOCS link at the top of each forum page could help.

[Steiner49er]

31 minutes ago, trurl said:

Post a screenshot of Main - Array Devices and Main - Pool Devices

Sorry for the quick quoted reply, but I opened up the webui in Edge and it's populated, so I'm guessing some caching issue in Brave.  Here's the Array/pool shots!

 

[Frank1940]

Do you have this problems with the missing WD drives?

 

      https://forums.unraid.net/topic/73717-solutions-shucked-wd-8tb-easystore-33v-pin-issue/

 

Do either of those servers use RAID SATA controllers.  Unraid is happiest with SATA controllers that are running in the IT-Mode rather than any RAID mode.   (Some RAID controllers can be configured to run with Unraid but they often have operating problems--- some of which don't show until you make a hardware replacement.

Edited September 11, 2022 by Frank1940

[Steiner49er]

2 hours ago, Frank1940 said:

Do you have this problems with the missing WD drives?

 

      https://forums.unraid.net/topic/73717-solutions-shucked-wd-8tb-easystore-33v-pin-issue/

 

Do either of those servers use RAID SATA controllers.  Unraid is happiest with SATA controllers that are running in the IT-Mode rather than any RAID mode.   (Some RAID controllers can be configured to run with Unraid but they often have operating problems--- some of which don't show until you make a hardware replacement.

Yes, both servers have (had?) RAID array controllers in them that connect the MB to the hot swap backplane.  Both controllers though I had in HBA mode, so they pass through the disks directly, RAID functions are disabled in HBA mode, which may be a contributing factor here, but I think this is more a matter that the disks are identifying differently (with more information it seems?) than they were on the last controller.

[Frank1940]

9 minutes ago, Steiner49er said:

Yes, both servers have (had?) RAID array controllers in them that connect the MB to the hot swap backplane.  Both controllers though I had in HBA mode, so they pass through the disks directly, RAID functions are disabled in HBA mode, which may be a contributing factor here, but I think this is more a matter that the disks are identifying differently (with more information it seems?) than they were on the last controller.

That is one of the things that can happen.  I believe most people have been able to figure out which disk is which because the actual serial number of the disk is used in the identifier string that the RAID controller generates.  It becomes a bit more difficult when you replace a RAID controller with another RAID controller as you are now dealing with two identifier strings!

Edited September 12, 2022 by Frank1940

[JorgeB]

6 hours ago, Steiner49er said:

Both controllers though I had in HBA mode, so they pass through the disks directly

RAID controller is still changing the devices names, and that is the problem, e.g. parity it was:

 

WDC_WD80EFAX-68LHPN0_7SGMN7JC

 

and now is :

WDC_WD80EFAX-68L_7SGMN7JC_35000cca252c8ef0a

 

You can do a new config but will have issues again if you change the controller in the future, we recommend using a true LSI HBA.

 

 

 

 

[Steiner49er]

5 hours ago, JorgeB said:

RAID controller is still changing the devices names, and that is the problem, e.g. parity it was:

 

WDC_WD80EFAX-68LHPN0_7SGMN7JC

 

and now is :

WDC_WD80EFAX-68L_7SGMN7JC_35000cca252c8ef0a

 

You can do a new config but will have issues again if you change the controller in the future, we recommend using a true LSI HBA.

 

 

 

 

I'm hoping I won't have to deal with that for a long time, but I'll be taking extra steps I think to try to mark what's what.

 

So in looking at the drives, I can tell which ones were the parity drives, and which 3 WD's were the data drives, the 1.2-1.8TB SAS dries didn't have any data on them yet since I had just recently brought up the Dell server and thankfully nothing had hit those drives yet.  With the parity drives though I can't tell which was parity 1 and which was parity 2, does that matter, is there a way for me to determine that?  Unassigned devices plugin helps to see that clearly with the parity/data, but luckily the two WD drives are WD80EFAX drives, so easy to tell apart from the data drives which are all WD80EMAZ... however, if I make a new config, does the order of the data drives matter??  I'm really hoping to (of course) not lose any data/backups/pictures, etc.

 

 

[JorgeB]

Parity:

WDC_WD80EFAX-68LHPN0_7SGMN7JC

Parity2:

WDC_WD80EFAX-68LHPN0_7SGM2J7C

 

Serial is similar but different.

[Steiner49er]

24 minutes ago, JorgeB said:

Parity:

WDC_WD80EFAX-68LHPN0_7SGMN7JC

Parity2:

WDC_WD80EFAX-68LHPN0_7SGM2J7C

 

Serial is similar but different.

Ah, I didn't spot it!  Thank you!!

 

Though in the end I suppose it didn't really matter.  I ran a new config and I have all my shares back and the parity disks are syncing now.

 

Thanks everyone!

[JorgeB]

13 minutes ago, Steiner49er said:

Though in the end I suppose it didn't really matter.

You could have checked "parity is already valid" to avoid the parity sync, and the array would have been protected from the start.

[Steiner49er]

22 minutes ago, JorgeB said:

You could have checked "parity is already valid" to avoid the parity sync, and the array would have been protected from the start.

Hmm, I missed the option... didn't see it I think, I would have checked that box.

[itimpi]

50 minutes ago, JorgeB said:

You could have checked "parity is already valid" to avoid the parity sync, and the array would have been protected from the start.

That would only work for parity2 if the disks were in the same order.   Not clear if this was the case.

[JorgeB]

18 minutes ago, itimpi said:

Not clear if this was the case.

Yes, good point, I assume yes.