Hello, I think all our advice would be to update to a later version of
the monitor-core, and preferably wait for the upcoming 2.5.0 version.
Even in the latest 2.4.0 release, there are ways to kill gmond by
querying its XML tree. We have tested the 2.5.0 extensively, and have no
reports of it pegging the cpu.
I am not aware of your problem, however none of our systems have run the
2.2.x version in a very long time.
Federico
On Tuesday, September 17, 2002, at 05:52 AM, Steven A. DuChene wrote:
Hello:
I am the person who is packaging ganglia as part of the OSCAR toolsuite.
We are currently using 2.2.3 of the ganglia-monitor-core package.
The only change to the package I had to make was to alter the spec file
so gmond wouldn't be started while the rpm was installed. The code is
as distributed from the ganglia sourceforge project.
A couple of our developers have reported the following problem.
Has anyone here ever seen or heard of this problem with 2.2.3?
I was considering updating to 2.4.1 but would like to have some
idea if this will have an effect on this problem.
----- Forwarded message from [EMAIL PROTECTED] -----
Subject: [Oscar-devel] [ oscar-Bugs-602940 ] gmond eats all cpu cycles
Date: Fri, 13 Sep 2002 08:26:25 -0700
Bugs item #602940, was opened at 2002-08-31 17:43
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=109368&aid=602940&group_id=
9368
Category: Packages
Group: 1.4
Status: Open
Resolution: None
Priority: 7
Submitted By: Jeff Squyres (jsquyres)
Assigned to: Steven A. DuChene (sad)
Summary: gmond eats all cpu cycles
Initial Comment:
When I reboot my nodes after a successful install
(i.e., they just downloaded and installed the image),
gmond is going crazy on the nodes.
uptime reports a load of over 3 for each machine., Top
shows that gmond is eating up 98% of the CPU (two gmond
threads).
"service gmond restart" on the nodes seems to fix this,
but this shouldn't be happening. This is from a fresh
reboot, after all (and the head node was up long before
the clients were).
----------------------------------------------------------------------
Comment By: Brian W. Barrett (bwbarrett)
Date: 2002-09-10 00:11
Message:
Logged In: YES
user_id=44328
I just saw the same thing on my RH 7.3 machine. The gmond
on the head node was eating up all the cycles on one of the
CPUs. Just one thread / process / whatever they are.
Restarting the gmond service seems to have fixed the
problem. Unfortunately, I didn't think to attach a debugger
or anything like that to see where the failure was occuring.
Next time...
----- End forwarded message -----
--
Steven A. DuChene [EMAIL PROTECTED]
[EMAIL PROTECTED]
http://www.mindspring.com/~sduchene/
-------------------------------------------------------
This SF.NET email is sponsored by: AMD - Your access to the experts
on Hammer Technology! Open Source & Linux Developers, register now
for the AMD Developer Symposium. Code: EX8664
http://www.developwithamd.com/developerlab
_______________________________________________
Ganglia-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ganglia-developers
Federico
Rocks Cluster Group, Camp X-Ray, SDSC, San Diego
GPG Fingerprint: 3C5E 47E7 BDF8 C14E ED92 92BB BA86 B2E6 0390 8845
