Hi,

We have a Dell Poweredge 1750 server with 2 harddrives in RAID 1 running Debian Etch. Yesterday one of the hardrives stopped working but no error message appeared in syslog. The only indication was the amber LED diode. Is there a software way how to check that both disks are OK?

Thanks for helping.

check out the hdparm and dmraid man pages to see if those will help you.

Thanks, but neither hdparm nor dmraid could find any information. It seems RAID to be completely hidden to the system.

If they are hardware RAID you might have to reboot into the RAID configuration utility and check the disks from there.

Yea, it sounds like you have a "real" hardware RAID then. That's great!

You'll probably have to take the system offline, boot it up, and play aroudn with the verification utilities offered by the RAID controller's bios / RAID management utility. (Loads after system BIOS POST.)

EDIT: (Oops was posting this while ncsuapex was simultaneously. "What he said.")

Dell has always been a big fan of LSI cards... most Linux hardware raid card drivers are actually dumping relatively easy to read statistics to somewhere in proc, sometimes sys, but there's no API for it, so every driver tends to be different. Most of the high end ones have software packages (not exactly open ones, so you'll have a hard time finding them in Deb), that you can install and either manipulate the card live, or set up a cron job to period checks, diffs and reports. Again, it all comes down to what kind of card. /sbin/lspci will tell you that. Post back here and I might be able to tell you more.