HDD goes read-only. Funky error output

[Tuork]

I'm working on a linux cluster at my Uni and one of the nodes keeps going bonkers on me every so often. As far as I can tell, there's some error with the HDD and it turns read-only. When I looked at the output coming from the screen it gave me this (wrote it down as login doesn't work):

----------------------------------------
ata2.00: status: { DRDY }
ata2.00: exception Emask 0x0 Sact 0x0 SErr 0x0 actiona 0x2 frozen
ata2.00: cmd 25/00:20:ef:7f:31/00:00:2x:00:00/e0 tag 0 dma 16384 in
res 40/00:01:00:4f:ce/00:00:00:00:00/00 Emask 0x4 (timeout)

end_request: I/O errpr, dev sdb, sector 741441519


Any idea what the hell is going on? This has to be at least the third time it's happend.

Many thanks.

[JonArbuckle]

I'm working on a linux cluster at my Uni and one of the nodes keeps going bonkers on me every so often. As far as I can tell, there's some error with the HDD and it turns read-only. When I looked at the output coming from the screen it gave me this (wrote it down as login doesn't work):

----------------------------------------
ata2.00: status: { DRDY }
ata2.00: exception Emask 0x0 Sact 0x0 SErr 0x0 actiona 0x2 frozen
ata2.00: cmd 25/00:20:ef:7f:31/00:00:2x:00:00/e0 tag 0 dma 16384 in
res 40/00:01:00:4f:ce/00:00:00:00:00/00 Emask 0x4 (timeout)

end_request: I/O errpr, dev sdb, sector 741441519


Any idea what the hell is going on? This has to be at least the third time it's happend.

Many thanks.

Time to back up data...
You can use SMART to check the drive status... but it's probably dying...

[Tuork]

I was looking into the /var/log/messages and the last line is always:

smartd[3623]: smartd has fork()ed into background mode. New PID=3623.


Any ideas? My guess is smartd is ****ing everything up.. though it doesn't make much sense seeing as it's supposed to monitor the disk...

I tried setting smartd to tell me whenever something goes wrong (via mail), but I don't think I did it right. Can anyone tell me how to do that?

Many thanks

[Dizzle7677]

Here's a relevant thread from a while ago about Smartd and the errors.
http://www.mail-archive.com/linux-ide@vger.kernel.org/msg11248.html

[Tuork]

Thanks mate. It says i should update smartd. I'll give that a go and see what happens.