Disk5 in Error State

[TexasTitan915]

I've been having random issues with a particular drive in my server. Disk5 will work perfectly fine one day, then after powering down the server for some time and powering it back up, it'll boot up with the drive in an error state. No read errors or anything. The first time it happened I powered down the array, removed the disk, started the array back up and then powered down and reconnected the drive. Powered back up, assigned the disk to Disk5 and rebuilt the data onto the drive. No issues while the server was up for about a week. I powered down, as part of routine maintenance, and here we are again after booting it back up. I've attached a copy of the syslog.

I see instances of "Synchronize Cache" failing on sdb, then sdc, then sdd. Could this be an instance where the power supply is going bad? It's an older 500w Corsair CX500.

hurtadoserver-syslog-20200722-2040.zip

[TexasTitan915]

I swapped the power supply just to rule that out and i'm pre-clearing the drive to rule out any issues with the drive itself.

[Gragorg]

Run an EXTENDED SMART test on the drive if you haven't already.

[TexasTitan915]

On 7/22/2020 at 9:39 PM, Gragorg said:

Run an EXTENDED SMART test on the drive if you haven't already.

I ran two preclear cycles and all was good. Currently running an extended SMART test, will post back with results.

[TexasTitan915]

Extended SMART test finished without error. SMART report and current diagnostic reports attached. 

hurtadoserver-smart-20200728-0816.zip hurtadoserver-diagnostics-20200728-0819.zip

[JorgeB]

Disk dropped right after spin down:

 

Jul 22 14:34:43 HurtadoServer kernel: mdcmd (45): spindown 5
Jul 22 14:35:58 HurtadoServer kernel: sd 12:0:3:0: device_block, handle(0x000b)
Jul 22 14:36:00 HurtadoServer kernel: sd 12:0:3:0: device_unblock and setting to running, handle(0x000b)
Jul 22 14:36:00 HurtadoServer kernel: sd 12:0:3:0: [sdi] Synchronizing SCSI cache
Jul 22 14:36:00 HurtadoServer kernel: sd 12:0:3:0: [sdi] Synchronize Cache(10) failed: Result: hostbyte=0x01 driverbyte=0x00
Jul 22 14:36:00 HurtadoServer kernel: mpt2sas_cm0: removing handle(0x000b), sas_addr(0x4433221105000000)
Jul 22 14:36:00 HurtadoServer kernel: mpt2sas_cm0: enclosure logical id(0x51866da06d30c200), slot(6)

 

Set spin down to never on that drive and see if it makes a difference.