My gmond's always bloat quite large. To combat this, I've dedicated a
single host with massive amount of swap to run a gmonds as a
'collector' role. Every time I add a new cluster, I add a new gmond
instance on the collector box with a new port.
The recent discussion about setting dmax caused me
On Thu, Feb 23, 2012 at 11:06 AM, Matt Massie m...@massie.us wrote:
Each unique metric (keyed on metric name) requires memory space in gmond.
A good test is to peek at the number of metrics in gmond over time, e.g.
$ telnet localhost 8649 | grep METRIC | wc -l
If the number of metrics over
This module is neat, but it's never been a viable module to run in
production in any site I've worked at. I disable it when I detect it
with config management, but if it happens to slip through, cpu on the
host is driven way up with the excessive netstat calls.
Does anybody else notice this?
You choose a single gmond in each cluster as a collector. Gmetad only
polls the collector for each cluster. In unicast mode, you set all
your gmond nodes to send an XDR formatted UDP datagram to the
collectors. In multicast mode, you can choose the collector at random
from gmetad's perspective.
Is there a limit on number of metrics a host can send?
We've got a gmond instance that we've loaded 3700 metrics. When run
in debug mode, gmond reports that it sent with 0 errors, but the
receiving gmond (we're running unicast mode) only shows 224 metrics in
the XML structure when we query it.
On Thu, Dec 9, 2010 at 2:59 PM, Stephen Tunney stun...@tsavo.com wrote:
Greetings everyone,
I am attempting to install ganglia 3.1.7 onto CentOS 5.5 to run alongside
Hadoop. I was to run 3.1.X to get support for python modules so that I can
in turn get email alerts from our cluster.
I
You need modpython.conf to include any .pyconf's not gmond.conf
itself. Change the include mysqld.pyconf to:
include ('/etc/ganglia/conf.d/*.conf')
then in /etc/ganglia/conf.d/modpython.conf you should have yet another include:
grep conf /etc/ganglia/conf.d/modpython.conf
include
On Tue, Dec 7, 2010 at 5:56 AM, Antonio Óscar Balmaseda
antonio.o.balmas...@gmail.com wrote:
Hello, again,
Alex, thanks for your response. I rechecked all the permissions and
everything was fine.
2010/12/6 David Birdsong david.birds...@gmail.com
If you put that open inside of metric_init
If you put that open inside of metric_init, start gmond as root, then
the filehandle will be created before gmond drops it's privileges to
the user configured.
On Mon, Dec 6, 2010 at 4:16 PM, Antonio Óscar Balmaseda
antonio.o.balmas...@gmail.com wrote:
Hi, everyone,
I have a strange problem.
On Fri, Nov 12, 2010 at 11:34 PM, Martin Knoblauch kn...@knobisoft.de wrote:
- Original Message
From: David Birdsong david.birds...@gmail.com
To: Whit Blauvelt w...@transpect.com
Cc: Martin Knoblauch kn...@knobisoft.de;
ganglia-general@lists.sourceforge.net
Sent: Fri, November 12
Hey all, I've cleaned up some of the functions and classes that I've
been reusing while writing various gmond python metrics. I want to
put them in the ganglia_contrib package on github. I forked it and
added my package via submodule -submodule sound ok?
Also, before I add it or do the pull
I'm not there anymore, but I think it was 3.1.2.
On Thu, Oct 14, 2010 at 4:23 PM, Anton Yurchenko ayurche...@gmail.com wrote:
Well that is good to know :)
What version of ganglia are you running?
Thanks!
On 10/14/2010 4:21 PM, David Birdsong wrote:
FYI, we did exactly this for ~4-5
On Tue, Sep 21, 2010 at 11:33 AM, Stevens, Weston J
weston.j.stev...@boeing.com wrote:
Sure does thanks! I guess dig into the source code of the DSO modules to make
it not recognize its own cycles? Bad idea?
Obviously, if Ganglia is using up considerable resources, one would want to
know.
Anybody have some tips as to why gmond is converting my float metric
that starts out looking like this:
fe_nginx.http_req_time 1.133985
which was collected by adding this to my get_metric callback and
running gmond in debug mode:
print sys.stderr, 'fe_nginx.http_req_time %00f' % v
On Thu, May 6, 2010 at 3:24 PM, Bernard Li bern...@vanhpc.org wrote:
Hi David:
On Thu, May 6, 2010 at 2:24 PM, David Birdsong david.birds...@gmail.com
wrote:
here's the relevant descriptor info:
desc_template = {
'call_back': self.get_metric,
'time_max
On Fri, Dec 4, 2009 at 10:35 AM, Cassandra Pugh cp...@pppl.gov wrote:
Hello,
I have ganglia 3.1.2 running on RHEL 32bit.
I have often noticed that the graphs have gaps in them, and am asking for
help on how to eliminate these gaps.
If I happen to access the ganglia page while a gap is
I have a few custom counter metrics that when restarted show a spike
so high on the ganglia we graphs that the metric is rendered useless
until enough time passes that the spike is outside of the time window.
My guess is that gmond reads in the counter values and derives a rate
so the first value
On Tue, Oct 6, 2009 at 10:29 AM, Brad Nicholes bnicho...@novell.com wrote:
On 10/5/2009 at 9:54 PM, in message
dcccdf790910052054p47873660h56ea883035f98...@mail.gmail.com, David Birdsong
david.birds...@gmail.com wrote:
say this is my *.pyconf
i'd like to be able to send a list through
say this is my *.pyconf
i'd like to be able to send a list through the config file to the
paramateres that are passed at init time for my python gmond module.
is there a syntax for allowing a list to be specified...something
like:
modules {
module {
name = lb-monitor
language = python
nice...we use ganglia here at imageshack.
On Sun, Aug 16, 2009 at 8:49 PM, Thanachote Pothanantthan...@th.ibm.com wrote:
Hi Seth,
Thank you for you reply and sorry if duplicate.
I just try for only 1 node (client, server and web front are all in the same
node).
About 'memory_limit' in
from a fedora rpm will call
'killproc' which is defined by the distro. killproc defaults to send
a SIGTERM first, wait then send a SIGKILL.
shouldn't gmond catch a SIGTERM and exit gracefully?
On Wed, May 27, 2009 at 1:40 PM, David Birdsong
david.birds...@gmail.com wrote:
thanks bernard, that sped
dcccdf790905232343y76481e5dw6c1df62bc732c...@mail.gmail.com, David Birdsong
david.birds...@gmail.com wrote:
I have a python module that spawns a separate thread that collects
data off of a pipe.
Everything runs fine, but I'm finding that metric_cleanup is never
called. When I strace the PID of the worker thread(in Linux
I have a python module that spawns a separate thread that collects
data off of a pipe.
Everything runs fine, but I'm finding that metric_cleanup is never
called. When I strace the PID of the worker thread(in Linux so it
get's it's own PID), I see it gets a SIGTERM when I stop gmond instead
of
, 2009 at 11:43 PM, David Birdsong
david.birds...@gmail.com wrote:
I have a python module that spawns a separate thread that collects
data off of a pipe.
Everything runs fine, but I'm finding that metric_cleanup is never
called. When I strace the PID of the worker thread(in Linux so it
get's it's
Yeah, thanks. I've been wondering about this functionality too as we
overload our boxes to do very different things. A separate cluster
view for each function -complete with summaries, is very useful.
On Thu, Jan 22, 2009 at 5:03 PM, Brad Fino b...@rockyou.com wrote:
I've run into this same
I'm having the same exact problem. I can run the if __name__ ==
'__main__' test and have the script define and execute the callbacks
via the descriptor list, but when running from gmond, all that gets
called is the metric_init()
I put an open in the init and set the open filename to global as a
that the user that gmond is running as, has sufficient permissions to perform
any work being done by the modules.
Brad
On 12/12/2008 at 4:24 AM, in message
dcccdf790812120324j79fbd1e1y14ebf50f6313f...@mail.gmail.com, David
Birdsong
david.birds...@gmail.com wrote:
I'm having the same exact
27 matches
Mail list logo