Hello,
A follow up to my previous email. Turns out that there was an issue
with awips2 messages getting erroneously sent to /var/log/messages.
This caused the 2GB /var partition to fill prematurely. I modified
/etc/rsyslog.conf as follows:
# Log anything (except mail) of level info or higher.
# Don't log private authentication messages!
*.info;mail.none;authpriv.none;cron.none;local0.none
/var/log/messages
local0.none /var/log/messages
I removed the large log files & rebooted the system. The 15 min CPU
load has not gotten higher than 2.11 in the last two days which is much
more acceptable.
Happy Friday!
Brendon
Brendon Hoch wrote:
Hello,
We have our EDEX server running with the default Unidata configuration
on a server that is a few years old, but should theoretically be beefy
enough to run things smoothly. The server has 8 CPUs (Xeon X5450 @
3.00GHz, 6144 KB cache) and 16 GB of memory. We are running version
14.4.1-4n6 under CentOS 6.7.
After running for over two weeks with little to no external requests
from CAVE clients, the CPU load is consistently high, with 15 min load
average about 5-6 and regularly getting to levels of 10-15. The number
of threads has been hovering near 2000 since we have started running EDEX.
Is this type of performance normal? Are there changes that can be made
to sysctl.conf or other configuration files to potentially reduce the
CPU load & threads?
Any help is greatly appreciated.
Thanks,
Brendon
--
Brendon Hoch, M.S.
Director of IT & Instrumentation
Judd Gregg Meteorology Institute
MSC 48, Boyd Hall 319
Plymouth State University
Plymouth, NH 03264
(603)535-2818 Fax: (603)535-2723
