Re: [tlug] Hi from new server and a couple of questions

[scott <scott@example.com>]

Hi everybody,

First of all thanks for the tips so far. I think I'll resolve the mail 
issue by using IMAP and forget about trying to chown the mailbox.

On Tue, 11 Jul 2006, GMO Unix Erin D. Hughes wrote:

> I may have missed it before but some details would be helpful like what
> distro are you running what kernel and what base packages/src's have you
> installed. I saw a little further down the thread some one said your
> information was unhelpful and I have to agree it is rather incomplete.

Sorry about that- Hardware is dual core AMD opteron 64 bit on Tyan 
server board, with boot partition on ide /dev/hda1. The rest of the 
filesystem is somewhat complex, I have 2 software raid 5 arrays running on 
3 SATA drives. Basically I split each sata into 2 linux raid partitions, 
then made 2 software raid5 devices. They look like this:

md1 : active raid5 sdc2[2] sdb2[1] sda2[0]
      464486912 blocks level 5, 256k chunk, algorithm 2 [3/3] [UUU]

md0 : active raid5 sdc1[2] sdb1[1] sda1[0]
      160649728 blocks level 5, 256k chunk, algorithm 2 [3/3] [UUU]

On each of these I have created a physical volume:

  PV /dev/md1   VG VolGroupFiles   lvm2 [442.97 GB / 101.16 GB free]
  PV /dev/md0   VG VolGroupRoot    lvm2 [153.19 GB / 32.00 MB free]

and created the following logical volumes:

  ACTIVE            '/dev/VolGroupFiles/LogVolHome' [97.66 GB] inherit
  ACTIVE            '/dev/VolGroupFiles/LogVolTmp' [48.84 GB] inherit
  ACTIVE            '/dev/VolGroupFiles/LogVolUsr' [97.66 GB] inherit
  ACTIVE            '/dev/VolGroupFiles/LogVolVar' [97.66 GB] inherit
  ACTIVE            '/dev/VolGroupRoot/LogVolRoot' [153.16 GB] inherit

Initially I had the /tmp filesystem running on the above logical volume, 
but because of system freezing I migrated /tmp to a different sata drive 
/dev/sdd1, then did the chcon etc and updated fstab. So now filesystem 
looks like this:

[root@example.com scott]# mount
/dev/mapper/VolGroupRoot-LogVolRoot on / type ext3 (rw)
none on /proc type proc (rw)
none on /sys type sysfs (rw)
none on /dev/pts type devpts (rw,gid=5,mode=620)
usbfs on /proc/bus/usb type usbfs (rw)
/dev/mapper/pdc_edcffdbc1 on /boot type ext3 (rw)
none on /dev/shm type tmpfs (rw)
/dev/mapper/VolGroupFiles-LogVolHome on /home type ext3 (rw)
/dev/sdd1 on /tmp type ext2 (rw)
/dev/mapper/VolGroupFiles-LogVolUsr on /usr type ext3 (rw)
/dev/mapper/VolGroupFiles-LogVolVar on /var type ext3 (rw)
none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw)

The /dev/mapper/pdc_edcffdbc1 is the boot ide /dev/hda1

The OS is Centos4. Everything is working like a charm, including APC 
monitoring via apcupsd and temp monitoring via gkrellm, the only issue is 
the occasional freeze when using xine to watch a .wmv file or sometimes 
xmms crashes on .wav files.

> > Are there any selinux gurus in the
> > house and if so have you experienced this problem?

The only reason I was thinking the freeze was caused by selinux was 
because on reboot after each freeze, fsck tags the /tmp filesystem as 
dirty and asks for root password to perform maintenance. Also there are 
previous bug reports about how some programs crash because /tmp cannot be 
mounted 'noexec' with selinux enforcing. The logs don't contain any clues 
because the system freezes before anything can be written.

Botond mentioned using strace as a way to troubleshoot this- I've never 
used strace before- would I initiate an strace on the PID for example of 
xmms before it crashes? I've read the fine manual for strace but I need to 
do a lot more studying before I can grep this command.

Again, thanks for your advice everyone.

Cheers,

Scott VanDusen