RAID degraded - but no LED shows degraded HD

[rince]

Hi,

tonight, my NAS detected a problem on Disk 3 (I have all five HDs within one Raid5 so there is no data loss). I got an email telling me that the system is in degraded mode.
But when I checked the LEDs on the system, I saw that no LED except all 5 blue ones are running; no red light showing the failed disk.
Is this a bug or how could I check which hard drive physically has failed?

[omega]

Hi rince,

if you have SSH access to your box, try this:

mdadm -D /dev/md0

In the output you should see which drive is degraded.

Then do this command:

cat /proc/scsi/scsi

In the ouput you can find that drive in the "Thecus:" line withe the prefix "Disk:". The "Tray:" entry should then give you the Slot for that drive.

Hope that helps...

Andreas

EDIT: use mdadm instead of mdstat (mdstat was wrong!)

[rince]

Hi,

interesting - there is no mdstat on my machine (Firmware 1.00.07), but since it was disk 3 I knew (or guessed) already it is the one in the middle.
The question is still - why is there no red led? shouldn't there be a red LED telling me which hard drive is bad?

[omega]

Hi rince,

I'm sorry, the command is not mdstat but mdadm. I'll change my post so nobody else is doing the wrong command.

Maybe you want to give it a try again....

Andreas

[rince]

Hi,

yepp, now it has the correct answer - it is the one in the middle (I have to say, I like that output!).

So now I'll wait for my replacement drive and then just rebuild the raid

Thanks for your help!

[quick]

Q1: Was it possible to find the faulty drive on the RAID screen or the disk-screen in the Web-interface

I think it is important that such events show up in a status-screen on either the RAID or the disks. The Web-GUI and the LEDs are whats to play with on a standard box....

[omega]

rince

Nice that it finally worked for you....

But I really wonder why the LED on the drive was not indicating the failure....
thecus: are there any explainations for this?

quick

You're absolutely right! This kind of information should be available on the web pages....
Maybe we should add this information to the INFO module....

The current firmware doesn't show the faulty drive on the web pages. If the drive does have a erroneous SMART status the you'll see it on the Disks page. On the RAID page you only see if the RAID is healthy or not.
It would be quite easy to create a page which shows the real status of each drive.

thecus: you should really think about to show the RAID status for every disk including it's physical location. This is really a must for a good device.....

Andreas

[kevincy]

There have a command could display all action about DISK .

/img/bin/savelog /app/cfg/logfile

[jamis]

My array went degraded last night out of the blue.... though the yellow drive light went solid and the LCD showed a message stating that drive 4 was failed. It also beeped (I haven't got email working yet).

I unplugged and plugged the drive back in and it rebuilt the raid.

This was the third time that this has happened on various drives in the array.

[rinconmike]

Are these failed drives or an issue with the 5200?

From the last post, it sounds line the 5200 thought the drive was bad, but unplugging it and plugging it back in fixed the issue. Is that correct?

Also, if they are failed drives, how old are the drives and what models are they?

thanks

[jamis]

I don't think they are failed drives. I think there is something else going on, but I'm not sure exactly what. I'll need to poke into the logs and see if I can find anything. There are no SMART errors on the drive that went down last night and like I said, it rebuilt fine after i plugged the same drive back in.

This is the third time total this has happened... if I remember correctly, on a different drive each time.