asked on

RedHat Linux 9 - Trouble with AMD

I have a cluster of Linux machines running RedHat 9 (2.4.20)
All machines have their amd service configured so as all machines can access the others disks.
My problem is that at times the amd process takes up >50% cpu on ALL machines!!!
Does anyone have a clue?

wesly_chen

Hi,

amd is the automount daemon. As my experience, amd doesn't work well under multiple heavy data transfer through (/net).
So for the NFS server with frequenty data access, you better use hard mount (specify in /etc/fstab or mount in /etc/rc.d/rc.local ) so you can specify the mount option to tune the NFS performance.

Cross NFS mounting is the CPU and network killer. Besides it is the trouble maker even though it is convenient by use of amd or autofs.

Tuning NFS/amd/autofs is the job for Unix systems adminstrators..... ^_^

Good luck,

Wesly

shayb022800

ASKER

Hi Wesly,
Thanks for the reply. Althogh I personally am not a Unix Sys Admin, all you have said above is known to me. And yet, my Unix Sys Admins have trouble locating and solving the problem of AMD taking > 50% of the CPU.
If I am not mistaken once the mount is established (automatically) minimal resources are needed as long as the mount is active.
So again what could be the problem? what should I ask my Sys Admins to check/do?

ASKER CERTIFIED SOLUTION

wesly_chen

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

shayb022800

ASKER

Wesly,
I will pass this info to our sys admins and let you know how it goes.