Running RH9. I have a PCI IDE controller with two 80GB harddrives attached, running as a RAID 1 array. One drive is the primary master, the other is the secondary master. One of the drives has developed a problem.

The RAID array is used only for file storage. I have a separate 40 GB HD on a motherboard IDE controller which has the OS loaded on it.

My questions are:
1. What utilities can I run to check each drive? How would I do it? Unmount, then ... , etc.
2. I would like to make sure that the problem is not with the PCI controller, so I would like to put both drives on the secondary channel of the motherboard controller as master and slave. What would I need to do to accomplish this?
3. If I have to replace the failed drive, what would I need to do? Can I just put in the new drive and the OS will take care of it?

I am a newbie at this stuff, and I setup the server about a year ago. I have forgotten many of the things I did to setup the RAID array. So thanks in advance for any help you can provide!

Frank

Are you using md raid, or some non-native software raid (eg. promise software raid or similar)?

Must be md raid, since I didn't load anything special. The array is referenced as md0.

I think this is the HOWTO you want to read.
http://www.tldp.org/HOWTO/Software-RAID-HOWTO.html

A couple of excerpts to answer your question.
Take a look at /proc/mdstat.

Finally, remember that you can always use raidtools or mdadm to check the arrays out.
mdadm --detail /dev/mdx
lsraid -a /dev/mdx

If you are using a 2.6 kernel, some of the older documentation may be out of date, as some things like array rebuilding are now done in the kernel.

Thanks for the help jschiwal!

I don't seem to have either raidtools or mdadm. I am reading through the how-to.