About a week or so ago I bought a couple of 12tb drives to add to my array.  In the process I replaced a couple of Syba 1xpci-e 2x SATA controllers with an LSI SAS9201 8i H3-25268 SAS card.

 

Everything installed ok, and my array was up and running again.  I added the unassigned disk plugin, and the preclear add on and I started to preclear the 12tb drives.

 

Again, everything looked good.  Until... Unraid went completely unresponsive.  I couldn't access it using its IP address.  I couldn't putty in to shut it down.  I usually run my server headless, but I plugged in a monitor and it didn't work either.

 

So I performed a hard reboot

Parity check took 8 hours but everything came back fine.

 

Saw something in the syslog that made me think I needed to update my mobo bios.

 

Updated the bios today.

 

Everything seemed to be working well so I resumed the preclear on those two drives.

 

Couple hours later and my server was unresponsive again.  I would appreciate whatever help you can provide.  I have attached the last three diagnositics (which I found on the unraid flashdrive) as well as the most recent syslog.

 

 

tower-diagnostics-20241110-1726 (1).zip tower-diagnostics-20241110-1517 (1).zip tower-diagnostics-20241106-1332 (1).zip tower-syslog-20241110-2327.zip

Enable the syslog server and post that after a crash.

Have not had a crash again yet, but my unassigned devices are acting weird.

 

Currently on the "main" page, under unassigned devices it says: "Timeout: Failed to get unassigned devices information - check the syslog."

 

On the "Dashboard" page, the unassigned devices are there - or at least they were... now it only shows one of the unassigned devises. (and in the time it has taken me to write this post they are now both gone).

 

My CPU, an Intel® Core™ i5-3475S CPU @ 2.90GHz, is currently fluctuating between 85% - 100%

 

Upon checking the syslog, I'm not seeing any errors beyond the warnings that my ReiserFS disks are going to be removed from future kernels.

 

Is this a problem with my LSI card?  Or one of my breakout cables?  (Come to think of it, the new drives that are struggling to preclear are both on the same breakout cable with nothing else attached to that breakout cable).

 

Attached is the current syslog (mind you, the system is still up and running, er... limping along - the server is definitely less responsive with the CPU pinned as it is).

 

 

 

 

tower-syslog-20241111-1801.zip

There are what look like power/connection issues with two devices, sdi and sdg:

 

Nov 11 10:09:08 Tower kernel: sd 5:0:2:0: [sdi] tag#540 CDB: opcode=0x12 12 00 00 00 24 00
Nov 11 10:09:08 Tower kernel: scsi target5:0:2: handle(0x000b), sas_address(0x4433221103000000), phy(3)
Nov 11 10:09:08 Tower kernel: scsi target5:0:2: enclosure logical id(0x500605b0080c3560), slot(0) 
Nov 11 10:09:09 Tower kernel: sd 5:0:2:0: task abort: SUCCESS scmd(0x000000005cbc79a8)
Nov 11 10:09:09 Tower kernel: sd 5:0:0:0: attempting task abort!scmd(0x00000000ee17e36e), outstanding for 60348 ms & timeout 60000 ms
Nov 11 10:09:09 Tower kernel: sd 5:0:0:0: [sdg] tag#533 CDB: opcode=0x12 12 00 00 00 24 00
Nov 11 10:09:09 Tower kernel: scsi target5:0:0: handle(0x0009), sas_address(0x4433221101000000), phy(1)
Nov 11 10:09:09 Tower kernel: scsi target5:0:0: enclosure logical id(0x500605b0080c3560), slot(2) 
Nov 11 10:09:09 Tower kernel: sd 5:0:0:0: task abort: SUCCESS scmd(0x00000000ee17e36e)
Nov 11 10:09:09 Tower kernel: sd 5:0:0:0: Power-on or device reset occurred
Nov 11 10:09:09 Tower kernel: sd 5:0:0:0: Power-on or device reset occurred
Nov 11 10:09:11 Tower kernel: sd 5:0:2:0: Power-on or device reset occurred
Nov 11 10:10:51 Tower kernel: sd 5:0:0:0: attempting task abort!scmd(0x000000000473f1d1), outstanding for 30031 ms & timeout 30000 ms
Nov 11 10:10:51 Tower kernel: sd 5:0:0:0: [sdg] tag#567 CDB: opcode=0x12 12 00 00 00 24 00
Nov 11 10:10:51 Tower kernel: scsi target5:0:0: handle(0x0009), sas_address(0x4433221101000000), phy(1)
Nov 11 10:10:51 Tower kernel: scsi target5:0:0: enclosure logical id(0x500605b0080c3560), slot(2) 
Nov 11 10:10:52 Tower kernel: sd 5:0:0:0: task abort: SUCCESS scmd(0x000000000473f1d1)
Nov 11 10:10:52 Tower kernel: sd 5:0:0:0: Power-on or device reset occurred
Nov 11 10:10:52 Tower kernel: sd 5:0:0:0: Power-on or device reset occurred
Nov 11 10:10:59 Tower kernel: sd 5:0:0:0: attempting task abort!scmd(0x00000000d3b84950), outstanding for 7043 ms & timeout 7000 ms
Nov 11 10:10:59 Tower kernel: sd 5:0:0:0: [sdg] tag#538 CDB: opcode=0x12 12 00 00 00 fe 00
Nov 11 10:10:59 Tower kernel: scsi target5:0:0: handle(0x0009), sas_address(0x4433221101000000), phy(1)
Nov 11 10:10:59 Tower kernel: scsi target5:0:0: enclosure logical id(0x500605b0080c3560), slot(2) 
Nov 11 10:10:59 Tower kernel: sd 5:0:0:0: task abort: SUCCESS scmd(0x00000000d3b84950)
Nov 11 10:10:59 Tower kernel: sd 5:0:0:0: Power-on or device reset occurred
Nov 11 10:10:59 Tower kernel: sd 5:0:0:0: Power-on or device reset occurred

 

Check cables

I have swapped out the SAS breakout cable that was attached to the new drives.  We will see what happens.

This is getting really really frustrating.

 

As I noted above, even after swapping out the SAS breakout cable, I'm still seeing the same issues (my new drives keep disconnecting during preclear).

 

The only thing I can think to do now is swap the sata cables so that the new drives are connected directly to the motherboard and old drives are connected to the SAS breakout cable.

 

But even if that gets my new drives up and running properly is it going to do so at the cost of my older drives' stability?

 

 

Switched my two new 12tb drives to the onboard SATA ports and everything seems to be stable.  Fingers crossed.

is the lsi hba in it mode?

On 11/11/2024 at 8:42 PM, bmartino1 said:

is the lsi hba in it mode?

I think so.  Is there any way for me to check this?

quick at boot screen will show something...

If the BIOS screen flashes up as MegaRAID, then it's in RAID mode. It it shows HBA and has no mention of virtual drives, it's in HBA (IT) mode.

Unraid needs the HBA in IT mode for its Software raid to work correctly.
Example:
https://www.coryfiala.com/flashing-lsi-9211-8i-hba-to-it-mode/

It's also visible in the diags, but forum attachments are not working today.