Hello support,

 

after upgrade after upgrade to release 6.10.2 and try 6.10.3 the disk of parity go in state disabled. i have try to rollback but the issue is always present.

My server is HPE ProLiant MicroServer Gen10 and before this release all is ok after upgrade the system go to block one time at the day.

i have try to stop array and remove the disk of parity and adding after the start of array but the issue persist.

At attached i have collect the diagnostic.

 

Thanks in advance

tower-diagnostics-20220615-2206.zip

Have you followed the procedure documented here in the online documentations accessible via the ‘Manual’ link at the bottom of the GUI? 

Why do you have parity2 instead of parity? Not technically wrong just wondering.

 

Parity2 is slightly more complicated and won't let you change the order of disks without rebuilding.

After dinstall the disk of parity and Check the system and reinstall the disk of parity and check again now is all Ok.

Now upgrade to last release 6.10.3 reboot the system is OK.

 

After one day the issue is at all HD:

with this errore drive:

Jun 19 10:32:01 Tower kernel: sd 10:0:2:0: [sdd] tag#414 UNKNOWN(0x2003) Result: hostbyte=0x01 driverbyte=DRIVER_OK cmd_age=5s
Jun 19 10:32:01 Tower kernel: sd 10:0:2:0: [sdd] tag#414 CDB: opcode=0x8a 8a 00 00 00 00 02 24 c8 d6 d0 00 00 01 f8 00 00
Jun 19 10:32:01 Tower kernel: blk_update_request: I/O error, dev sdd, sector 9207076560 op 0x1:(WRITE) flags 0x0 phys_seg 63 prio class 0
Jun 19 10:32:01 Tower kernel: sd 10:0:2:0: [sdd] tag#415 UNKNOWN(0x2003) Result: hostbyte=0x01 driverbyte=DRIVER_OK cmd_age=5s
Jun 19 10:32:01 Tower kernel: sd 10:0:2:0: [sdd] tag#415 CDB: opcode=0x8a 8a 00 00 00 00 02 24 c8 d2 d0 00 00 04 00 00 00
Jun 19 10:32:01 Tower kernel: blk_update_request: I/O error, dev sdd, sector 9207075536 op 0x1:(WRITE) flags 0x4000 phys_seg 128 prio class 0
Jun 19 10:32:01 Tower kernel: sd 10:0:2:0: [sdd] Synchronizing SCSI cache
Jun 19 10:32:01 Tower kernel: sd 10:0:2:0: [sdd] Synchronize Cache(10) failed: Result: hostbyte=0x01 driverbyte=DRIVER_OK

 

i need to restore the 6.9.2 before the upgrade i dont have this issue its possible? some can help me?

Please post new diagnostics.

attached

tower-diagnostics-20220619-1243.zip

First update the LSI firmware to latest, all p20 release except the latest one have known issues, if errors continue post new diags before rebooting, diags you posted above are clean.

On 6/19/2022 at 12:50 PM, JorgeB said:

First update the LSI firmware to latest, all p20 release except the latest one have known issues, if errors continue post new diags before rebooting, diags you posted above are clean.

root@Tower:/home/Installer_P20_for_Linux/sas2flash_linux_i686_x86-64_rel# sas2flash -listall
LSI Corporation SAS2 Flash Utility
Version 20.00.00.00 (2014.09.18)
Copyright (c) 2008-2014 LSI Corporation. All rights reserved

        Adapter Selected is a LSI SAS: SAS2008(B2)

Num   Ctlr            FW Ver        NVDATA        x86-BIOS         PCI Addr
----------------------------------------------------------------------------

0  SAS2008(B2)     20.00.06.00    14.01.00.09    07.39.02.00     00:03:00:00

        Finished Processing Commands Successfully.
        Exiting SAS2Flash.
root@Tower:/home/Installer_P20_for_Linux/sas2flash_linux_i686_x86-64_rel#
 

 

But my version is 20 

tower-diagnostics-20220630-2357.zip

 

too much error the system have logs at 100

this is a error at tool-> system logs->

Fatal error: Allowed memory size of 134217728 bytes exhausted (tried to allocate 134115360 bytes) in /usr/local/emhttp/plugins/dynamix/include/Syslog.php on line 18

11 hours ago, pornoman79 said:

But my version is 20 

Yes, but it's not 20.00.07.00

 

On 6/19/2022 at 11:50 AM, JorgeB said:

all p20 release except the latest one have known issues

 

upgrade complete

 

root@Tower:/home/P20# sas2flash -listall
LSI Corporation SAS2 Flash Utility
Version 20.00.00.00 (2014.09.18)
Copyright (c) 2008-2014 LSI Corporation. All rights reserved

        Adapter Selected is a LSI SAS: SAS2008(B2)

Num   Ctlr            FW Ver        NVDATA        x86-BIOS         PCI Addr
----------------------------------------------------------------------------

0  SAS2008(B2)     20.00.07.00    14.01.00.09    07.39.02.00     00:03:00:00

        Finished Processing Commands Successfully.
        Exiting SAS2Flash.
root@Tower:/home/P20#

 

Rebuild on going 6 hr to complete crossing the finger
 

tower-diagnostics-20220701-1638.zip 

cross finger dont work 

again error on disk

ul  1 16:29:24 Tower kernel: blk_update_request: I/O error, dev sdb, sector 2526234808 op 0x0:(READ) flags 0x0 phys_seg 23 prio class 0
Jul  1 16:29:24 Tower kernel: blk_update_request: I/O error, dev sdb, sector 2526233784 op 0x0:(READ) flags 0x0 phys_seg 128 prio class 0
Jul  1 16:29:24 Tower kernel: blk_update_request: I/O error, dev sdb, sector 2526232760 op 0x0:(READ) flags 0x0 phys_seg 128 prio class 0
Jul  1 16:29:24 Tower kernel: blk_update_request: I/O error, dev sdb, sector 2526232752 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0
Jul  1 16:29:24 Tower kernel: blk_update_request: I/O error, dev sdb, sector 909136840 op 0x0:(READ) flags 0x0 phys_seg 32 prio class 0
Jul  1 16:29:24 Tower kernel: blk_update_request: I/O error, dev sdb, sector 2526030672 op 0x1:(WRITE) flags 0x0 phys_seg 128 prio class 0
Jul  1 16:29:24 Tower kernel: sd 10:0:0:0: [sdb] Synchronizing SCSI cache
Jul  1 16:29:24 Tower kernel: sd 10:0:0:0: [sdb] Synchronize Cache(10) failed: Result: hostbyte=0x01 driverbyte=DRIVER_OK

Jul  1 16:29:19 Tower kernel: mpt2sas_cm0: SAS host is non-operational !!!!

 

It's a problem with the HBA, it stops responding and Unraid loses contact with all disks, make sure it's sufficiently cooled, well seated and/or try a different PCIe slot if available, failing that try a different HBA.

jorgeB before upgrade of this release not issue is presente and for 5 month.

Can restore the last release Unraid 6.9.2?

 

Thanks in advance

Unlikely that is the problem, but you can, if you upgraded using the GUI there's a restore option (Tools -> Update OS)

i have done 2 upgrade 6.9.2 to 6.10.2 and again to 6.10.3 and dont find the release 6.9.2.

At the moment i have lost the array because i have 2 disk at fault.

Is it possible to go back to 6.9.2 manually and recover the array?

29 minutes ago, pornoman79 said:

i have done 2 upgrade 6.9.2 to 6.10.2 and again to 6.10.3 and dont find the release 6.9.2.

At the moment i have lost the array because i have 2 disk at fault.

Is it possible to go back to 6.9.2 manually and recover the array?

You may find this section of the online documentations accessible via the ‘Manual’ link at the bottom of the GUI to be of use?

On 7/1/2022 at 6:41 PM, itimpi said:

You may find this section of the online documentations accessible via the ‘Manual’ link at the bottom of the GUI to be of use?

i have restore the 6.9.2 but the array is offline because the parirty and other HD is a new device per the system.

can recovery it or definily lost?

attach new diagnostics to your NEXT post

5 minutes ago, pornoman79 said:

parirty and other HD is a new device

Did you keep your config folder when you downgraded?

Just now, trurl said:

Did you keep your config folder when you downgraded?

i have upload only file bz not other file is change the config folder is not present

tower-diagnostics-20220703-1911(1).zip

Do you know which disk is parity and which is disk1? We can get them from your earlier diagnostics if you don't

9 minutes ago, trurl said:

Do you know which disk is parity and which is disk1? We can get them from your earlier diagnostics if you don't

yes dev 1 is parity WDC_WD6003FFBX-68MU3N0_V9KGM6AL - 6 TB (sdb)

Your first diagnostics had parity2 instead of parity, but it looks like you must have rebuilt that as parity.

 

According to the diagnostics in this post

 

https://forums.unraid.net/topic/124911-parity-disabled-after-upgrade-6102/?do=findComment&comment=1139961

 

Disk1 was disabled, and parity was not, so technically, you should be attempting to rebuild disk1 now.

 

Do you have a spare disk to rebuild disk1? That would be simplest and safest, then you would still have original disk1 in case rebuild didn't work out.

 

Whatever you do it won't be as simple as rebuilding a disk as explained in the wiki since Unraid doesn't currently know what to do with parity and disk1, but we can work through that.