1

A year ago I was given an old IBM x336 server to run some personal projects on, now I knew nothing about servers at all so just put some more RAM in, wiped and disks and installed Ubuntu Server on it and have been using it to host some web services and to store data.

Recently on the front of the server the Diagnostic LEDs, drop-down light path diagnostics panel, has been red on the hard disk section. I assume either one of the disks has died or the (standard) Raid-1 array has broken - I believe this server uses Hardware RAID-1 by default.

Now as far as I am aware, remember I had no real understanding of servers when I set this up, I have no IBM specific software on the machine and I have no idea how to configure it remotely. I did notice there appears to be an IBM ethernet port? (As well as the two standard network adapters).

So really this is two questions:

  • How do I diagnose the red light and rebuild the RAID array if required?
    • If this is not possible how to install new drive(s)?
  • How do I set this server up properly with IBM tools etc so I can monitor it?

I usually just do everything via ssh and configure services, I am potentially prepared to wipe the server and start again properly following your advice if there is a way to store some of the data on it temporarily.

Thanks for your time,


Please note lots of places (and IBM's own manual) mention things like ServeRaid and other tools, I know nothing about these, just basic use of ubuntu server.

4
  • -1 for obvious lack of research. I don't even have to google "serveraid" and "IBM x336 hard disk red light" to know they will come back with the info you're after. Aug 26, 2011 at 9:34
  • @Ben: When I google "IBM x336 hard disk red light" I end up exactly here :)
    – Sven
    Aug 26, 2011 at 10:15
  • 1
    OK, I was being grumpy earlier, sorry about that. I was on mobile earlier, but now I'm on my laptop I have actually Googled the things I said you should Google, and surprisingly I didn't get much useful info. I hope to rectify that with my answer. I've also rescinded my downvote. Aug 26, 2011 at 11:39
  • @Ben no problem, thanks for the update, yes I did do a good google before I asked question, problem is answers are very rare, and if I find one I either don't understand the answer or it is not relavant!
    – Pez Cuckow
    Aug 26, 2011 at 13:42

1 Answer 1

4

ServeRAID is the software you use to configure your RAID arrays, and will also give you health information on your arrays and individual disks in your arrays. This is usually a bootable CD, and a GUI exists for Windows, however I'm not sure about Linux.

Red lights are usually bad, and often indicate a dead or dying disk. If you can't find a Linux version of ServeRAID to run while your box is still up, you'll have to reboot and run the bootable CD to see what's going on. The ServeRAID software should be able to identify which disk has died (or is dying) and should have the option to blink an LED on that specific disk to identify it. On server systems, individual hard disks usually have status indicator LEDs on them that illuminate when the disk is dead, along with another on the general server health panel.

As for fixing your problem, that depends on the RAID controller. Read its documentation, but if you're lucky it will be hot-swap and you can simply pull the dead one out and replace with a new disk of the same size & model, all while the server is still running.

You should always have a known good backup, just in case something bad happens, but decent server systems with decent RAID controllers are pretty solid when it comes to this kind of thing.

1
  • Thanks for the information, I will try and see if I can find a ServeRAID command line interface for Ubuntu Linux and see what I can work out. Else I guess it's a case of downloading and making a bootable CD, taking the server offline and running it from there.
    – Pez Cuckow
    Aug 28, 2011 at 12:43

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .