Re: [cobalt-users] HD failure cont'd

[Gerald Waugh <gwaugh@xxxxxxxxxxxxxxxxxxxxxxx>]

On Tuesday 18 June 2002 09:33 am, Martin wrote:
> Hi,
>
> A little while ago I asked for some help to find out what is wrong with my
> HDs in my RAQ4, since one of them failed according to the Cobalt
> Administration.
> One suggestion was to run the fsck command. I did, and there were some
> filesystem problems indeed. However, problems keep popping up (inode block
> bitmap differences, wrong free inodes count for groups). Is it normal to
> have those kind of problems all the time? I mean, when I fix those problems
> in the morning, in the afternoon they are there again...

No, you shiould not have all those problems.
I would take the drive out, install on a Linux desktop, and run test on it.
Or I would buy and install a new drive...
And just maybe I'd do both.

> I did some research in the Cobalt knowledge base, and found that using "cat
> /proc/mdstat" you should be able to find out which drive is messing up.
> Since I am new to the Unix filesystem, could anybody tell me what the
> following output means for me (can't find a man entry for mdstat)?
>
> Personalities : [linear] [raid0] [raid1] [raid5] [translucent]
> read_ahead 1024 sectors
> md1 : active raid1 hda1[0] 768000 blocks [2/1] [U_]
> md3 : active raid1 hda3[0] 205056 blocks [2/1] [U_]
> md4 : active raid1 hda4[0] 28879616 blocks [2/1] [U_]
> md6 : active raid1 hda6[0] 131456 blocks [2/1] [U_]
> unused devices: <none>

That is a normal (OK) mdstat

-- 
Gerald Waugh 
http://frontstreetnetworks.com  SOHO Networks & Web Site Hosting
Front Street Networks LLC     voice +1 203 785 0699 * fax +1 203 785 1787
229 Front Street, Ste. #C, New Haven CT 06513-3203