Today I have seen that one of my discs is lost.

 

 

I don't clearly see what had happened:

 

One disc with errors (Disc 5)

One disc lost (Disc 6)

One disc with errors (Disc 7) with one share that is not shown on my shares tab.
   I can see the share on MAIN tab on "browse /mnt/disc7" but there is no share on SHARES tab.

 

What must I do?

Attached diagnostics file.

 

Thankyou
Gus

 

 

unraid-media-diagnostics-20170421-1727.zip

Edited April 21, 20179 yr by zzgus

Do you have any disks connect on the Marvell controller (first 4 white ports)?

Quote

Do you have any disks connect on the Marvell controller (first 4 white ports)?

 

Don't tell me I have made the same mistake another time !!!

Some days ago I made changes to the disks on new cages but didn't remember had changed sata cables since last time.

Will check.

 

Do you see on the diagnostics file that those discs are connected to the marvel?

 

Thankyou
Gus

Edited April 21, 20179 yr by zzgus

Didn't check the logs before, still not sure since the beginning of the syslog is missing, but looks like a controller crashed, possibly the Intel SCU controller (the last 4 blue ports)

8 minutes ago, johnnie.black said:

Didn't check the logs before, still not sure since the beginning of the syslog is missing, but looks like a controller crashed, possibly the Intel SCU controller (the last 4 blue ports)

 

The MARVELL are problematic, the intel crash, buffff

Will a reboot solve that? how to proceed?

 

Thankyou
Gus

Best would be to reboot and post new diags.

2 hours ago, zzgus said:

the intel crash, buffff

 

Just to be clear about this, it's not the "normal" 6 port Intel controller, that is IMO the best controller anyone can use, very stable, but the 4 extra SCU ports:

 

05:00.0 Serial Attached SCSI controller [0107]: Intel Corporation C602 chipset 4-Port SATA Storage Control Unit [8086:1d6b] (rev 06)
    Subsystem: ASRock Incorporation C602 chipset 4-Port SATA Storage Control Unit [1849:1d6b]
    Kernel driver in use: isci
    Kernel modules: isci

 

These use a driver I've not much experience with but it they really were the problem then it's the first time I've seen this controller crash, unlike the almost daily issues I see in the forum with the Marvell controllers, especially with this series of Asrock boards where they are notoriously bad.

I have rebooted the system, changed sata ports and power but I continue to have thisc 6 miss.

Attached diagnostics.

 

As I don't use disc 7 (almost empty) whats the way to use it as disk 6 and recover it's contents from parity? is it possible?

 

Thankyou

Gus

unraid-media-diagnostics-20170421-2336.zip

All three affected disks are on the SCU ports, so I believe I was right and that controller's driver crashed, hopefully it was a one time thing.

 

Disk6 is not missing, it's just disable, unRAID disables the first disk it can't write to when a controller crashes (or the first 2 if using dual parity), disk looks fine so you should rebuild to the same disk, to do this you need to stop the array, unassign disk6, start the array, stop the array, reassign disk6 and start the array to begin the rebuild.

Edited April 21, 20179 yr by johnnie.black

As always @johnnie.blackmany thanks for your quick response.

 

Gus

 

@johnnie.black one last thing.

 

I have a spare single socket supermicro X9SRi-F motherboard.

http://www.supermicro.com/products/motherboard/Xeon/C600/X9SRi-F.cfm

As I no longer need the dual socket ASROCK motherboard I'm using now, do you think the supermicro will be a better option?

 

Thankyou
Gus

Difficult to say but I would guess that as long as the Asrock Marvell ports are not used stability should be similar.

@johnnie.black Well, mmmm not yet a month and another time some errors on disk 6 a before.

Attached diagnostics to see if there's something new.

I don't know what I will have to do with those "persistent" crashes :-(

 

Thankyou

Gus

 

unraid-media-diagnostics-20170515-2137.zip

Symptoms are the same, but happening again to the same disk may not be a coincidence, I would swap that disk with another from the main Intel controller, rebuild it, and in case it happens again you can see if the problem follows the disk or stays with the controller.

Edited May 15, 20179 yr by johnnie.black

@johnnie.blackbad news.

I did a rebuild with no luck, some read errors. (rebuild cancelled I think unraid showed on notifiactions tab)

I will remove this disc for safety.

 

Gus

If the rebuild failed on the main Intel controller it's probably a bad disk despite the healthy looking SMART.