Nestor wrote:
>..
> root 1043 0.0 0.5 12156 12152 ? SL 2005 0:00 mdmpd
I wouldn't necessarily accuse this of being _the problem_, but is there
any known reason for running mdmpd ?
from man mdmpd:
mdmpd - daemon to monitor MD multipath devices
are you _really_ running multipath devices???
I also note you are runnning some other services
portmap
vsftpd
httpd
whether-or-not they are suspect, you may wish to examine the active
connections
netstat -antpo
Is there anything funny-looking about the connections?
..and if you can tolerate dis-connections, perhaps try some experiments
of turning off (one-at-a-time) these services, and seeing what impact it
has on response.
Umm, perhaps you can grab your /var/log/messages.* for offline analysis
? are there any large number of recurring log entries
? disk errors
? scsi system errors
? ahc_dv (anybody know anything about this? is it adaptec, maybe google
for your rhel version, kernel-version and ahc_dv?)
I have seen ethernet modules (assuming it's a module, that is) get into
"funny" states. Again, if you can tolerate the disruption, maybe shut
down networking, remove and re-insmod the module, and restart the other
services.
You wouldn't think that networking would impact running processes like
dmesg, as you report, but, ...
..jim
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
