mount problem

[M. Tariq Ayub]

Hi,

I have two rp8400 in a clustered configuration. recently i created one mount point in one server that is not cluster aware and i have also entered its entry to /etc/fstab. but when i rebooted the system the file system was not mounted. What could be the reason?

[Rajeev Shukla]

One thing, If you reboot the NFS server then remember to export the filesystem.
But if just the NFS client is rebooted then the mount point should comeup by itself if correctly defined in /etc/fstab,

Are you able to mount it manually?
Have a look in /etc/rc.log do you see any error in nfs client startup

Also can we have a look at the entry you have put in /etc/fstab on nfs client (where you are mounting the filesystem)

[Mark Grant]

If this is a volume group that the cluster does not start up and /etc/lvmrc has the variable "AUTO_VG_ACTIVATE" set to 0, then you wil need to add the "vgchange -a y" commnad for this volume group to the custom_vg_activation() function within /etc/lvmrc. Otherwise the volume group will not be activated on boot and the file system can not be mounted.

[M. Tariq Ayub]

Hi,

The AUTO_VG_ACTIVATE is set to 0. and i need to activate that volume groum manually. now if i chage that variable to 1 then will it effect cluster environment. because at that time all the volume group of that server will be activated.

[Mark Grant]

Don't change it to 1. If you do, then your cluster volume groups will not be activated when you next boot the machine.

Look further down /etc/lvmrc and you'll see something like this.

custom_vg_activation()
{
# e.g. /sbin/vgchange -a y -s
# parallel_vg_sync "/dev/vg00 /dev/vg01"
# parallel_vg_sync "/dev/vg02 /dev/vg03"

/sbin/vgchange -a y /dev/vg00

From here you can see that "vg00" is activated from within this file. Just add a new entry under here for your new volume group.

[M. Tariq Ayub]

Hi,

i donot find a entry for vg00. i am attaching the lvmrc file. please look in to it.

[Mark Grant]

I am surprised vg00 isn't there. Just to confirm that we are on the right track though, look in /etc/rc.log. Somewhere in there I am expecting to see an error when trying to mount your file system saying something like the volume group has not been activated.

If this is the case just add the following to your /etc/lvmrc file in the section we have been looking at

vgchange -a y -s /dev/vgwhatever for the volume group containing your filesystem.

[M. Tariq Ayub]

Hi,

i found the entry in rc.log. it shows
mountall: cannot mount /dev/billprnvg/lvol01
mountall: diagnostics from mount
vxfs mount: Cannot open /dev/billprnvg/lvol01: No such device or address
checking quotas

It because of the vg is not activated.

Should i put the entry in
custom_vg_activation()
{
# e.g. /sbin/vgchange -a y -s
/sbin/vgchange -a y -s billprnvg
# parallel_vg_sync "/dev/vg00 /dev/vg01"
# parallel_vg_sync "/dev/vg02 /dev/vg03"
return 0
}

In the second line. My vg name is billprnvg

[Mark Grant]

Can you try these steps first (assuming that the filesystem is not already mounted)

This will confirm that the volume group is not already activated. You should get a message something like "volume group now activated"

vgchange -a y billprnvg

Now lets check that we can really mount the logical volume on a mount point

mount /dev/billprnvg /mnt

IF this works and you can see it mounted on /mnt in "bdf" then change /etc/lvmrc as discussed and as you have mentioned in your last post.

If the mount hasn't worked, I suspect that you haven't made the volume group correctly in the first place. But let me know how the above goes first.

[M. Tariq Ayub]

Hi,

I can mount manually. I will change the entry in that file