Problems 6.8.3...a real doozy

[MacModMachine]

Here is a doozy , been using unraid for 11 years now , never had to ask a question.....well today is that day.

 

Unraid 6.8.3

 

9 disks in array. Ryzen 3400G + 64GB Ram + LSI2008 controller.

 

Disks getting booted from array overnight. 

 

Replaced Controller

Replaced cabling + power supply (Data+power)

Replaced ram

Replaced motherboard + cpu (ryzen 3600 now)

replaced every disk with brand new sealed disks

 

ran memory test for 7 days with no issues. ran Freenas on same system with ZFS pool's and no issues for 14 days. 

 

still booting disks....in unraid.

 

grateful for any help as im going slightly crazy.....posting diag

fileserver-diagnostics-20200524-0740.zip

[Squid]

What do you mean by booted?

 

What I see is that the system looked like it started normally.  Then you unassigned the parity drives, reassigned them and restarted the array.  The system starts to build the parity information and then you cancelled it.

 

 

[MacModMachine]

9 minutes ago, Squid said:

What do you mean by booted?

 

What I see is that the system looked like it started normally.  Then you unassigned the parity drives, reassigned them and restarted the array.  The system starts to build the parity information and then you cancelled it.

thats my bad , i forgot to grab the diag before reboot. ill wait for it to show again and grab it again.

 

by booted , i mean they show a red X , however the disk remains fine. must be removed and readded to array. they are showing cannot write sector in the logs. however the disk is fine. scanned it several times with spinwrite. 

Quote

 

 

Edited May 24, 2020 by MacModMachine
forgot to answer question

[MacModMachine]

I forced the PCIE down to V2 from auto to see if that could possibly be the issue , since i have a H310 in this now. I have tried several brand new 9211's though.

[MacModMachine]

Had another drop out , grabbed the logs this time.

fileserver-diagnostics-20200526-2119.zip

[MacModMachine]

 

this is what i can see possibly being the issue, however not much information is given :

 

May 26 17:01:19 fileserver kernel: sas: --- Exit sas_scsi_recover_host: busy: 0 failed: 1 tries: 1
May 26 17:01:19 fileserver kernel: sd 10:0:1:0: [sdd] tag#504 UNKNOWN(0x2003) Result: hostbyte=0x04 driverbyte=0x00
May 26 17:01:19 fileserver kernel: sd 10:0:1:0: [sdd] tag#504 CDB: opcode=0x8a 8a 00 00 00 00 00 74 81 02 60 00 00 00 08 00 00
May 26 17:01:19 fileserver kernel: print_req_error: I/O error, dev sdd, sector 1954611808

 

 

[JorgeB]

On 5/24/2020 at 11:48 AM, MacModMachine said:

9 disks in array. Ryzen 3400G + 64GB Ram + LSI2008 controller.

You're also using a SASLP, can't see SMART since the disk dropped offline but those controllers are known to drop disks without a reason and not recommended for a long time.

[MacModMachine]

22 minutes ago, johnnie.black said:

You're also using a SASLP, can't see SMART since the disk dropped offline but those controllers are known to drop disks without a reason and not recommended for a long time.

That was added after , the disks in question are not on it. This problem started before that saslp was added. I can take it out it will make no difference with this problem.

[MacModMachine]

28 minutes ago, johnnie.black said:

You're also using a SASLP, can't see SMART since the disk dropped offline but those controllers are known to drop disks without a reason and not recommended for a long time.

crappp....ii think your right....i must have f'd up somewhere....the disk that failed was no doubt on that controller....ill have to hate myself for the rest of the day at minimum....

 

thanks...seriously...thanks.

 

ill remove that controller and burn it with a torch.