process commands very slow

Hello list,

We have a 32-node cluster.  I notice that sometimes every process-info 
command (e.g. (m)ps or (m)top) takes an extraordinary long time to 
execute.  This happens even under moderate load (a load average of 1 on 
each node).  Example:

$ time ps ax >/dev/null
real    3m18.055s
user    0m0.020s
sys     0m0.040s

Clearly, the CPU time is negligible, so it's not because the machine is 
under a too high load.  It does not seem to make a difference if ps or 
mps is used.

I assume it has something to do with the communication between the 
server and the diskless nodes, but even then 3 minutes is an awful long 
time.

Is there someone else who has seen this problem?  We are running Gentoo 
Linux with the official 2.4.26-om1 kernel.  We have a server + 32 
diskless nodes, all Pentium4 3GHz with 2GB of RAM (3GB on the server).


Cheers,
Jeroen Demeyer


-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click