On Mar 2, 2009, at 1:24 PM, Furnish, Trever G wrote:
Do your iowait percentages go up in top? Perhaps you have an
i/o-related problem.
No, %wa is 0%
Are your network cards running at the same duplex as your connected
switch ports?
Yes they are 1000 Full Duplex
Is kswapd busy (ie your system's thrashing) during the spikes?
No.
You might get useful records by installing the sysstat package and
then
configuring sar to run every minute, then using sar to report on the
collected data.
sysstat is installed, but the spikes seem to be happening so quickly I
can't get a read on it. For instance, "top" is running and all of the
sudden load average was 20, but cpu is 95% idle.
-----Original Message-----
From: redhat-list-bounces@xxxxxxxxxx [mailto:redhat-list-
bounces@xxxxxxxxxx] On Behalf Of Troy Knabe
Sent: Monday, March 02, 2009 4:05 PM
To: Red Hat
Subject: High Load and Low CPU & Memory
I have some brand new servers with 2 - Dual Core Xeons 3.33 Ghz and
48Gb of RAM. Intermittently they are spiking to high load averages
5+
or sometimes even 10+ and the application (oracle application server)
becomes unresponsive. We are moving them from 2 - Dual Core Opterons
2.4 Ghz and 12GB of RAM, and never experienced this type of
situation. Can anyone give me some suggestions on where to start
looking?
I am using Kerberos/LDAP authentication/authorization for the new
ones, but the ldap servers do not appear to be the issue. The oracle
user is local to each server.
Thanks
-Troy
--
redhat-list mailing list
unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe
https://www.redhat.com/mailman/listinfo/redhat-list
--
redhat-list mailing list
unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe
https://www.redhat.com/mailman/listinfo/redhat-list
--
redhat-list mailing list
unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe
https://www.redhat.com/mailman/listinfo/redhat-list
