On Tue, Sep 21, 2010 at 14:33, 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?
Good luck with that. :) I'm not aware of any monitoring tools that
explicitly track
their own impact" directly. For example, go run the "top" command on
an idle system. More likely than not, you'll see "top" running at the
"top" of the list (when sorting by CPU usage).
That said, in trunk, there is a mildly self-aware gmond module that
tracks several metrics related to gmond packets sent and received.
> Obviously, if Ganglia is using up considerable resources, one would want to
> know. However, the purpose of Ganglia is to measure the performance of
> everything else, Ganglia can skew this data. Ideally and naturally, one would
> like to measure performance without using up any resources in doing so.
Nothing is free. If you want metrics, you have to spend some time
collecting them.
In my experience, gmond doesn't use much in the way of CPU cycles.
Collecting metrics at sane intervals (multiple seconds, as opposed to
millisecond or less), costs almost nothing on modern CPUs. However,
an insane polling time ("collect everything, every second") will
probably have an adverse impact.
What I *have* seen is that gmond can use a lot (relatively speaking)
of RAM. I have *not* done any empiric tests with this, but I believe
that it scales with the number of metrics stored. Thus, if you use
the default multicast settings, where all nodes store all metrics for
all nodes, the memory usage can get large (you have X hosts, each with
Y metrics, stored on each host...).
You can avoid this by using unicast to specifically designated
"collector" gmonds (then having gmetad poll those for overall status).
> -----Original Message-----
> From: Martin Knoblauch [mailto:kn...@knobisoft.de]
> Sent: Tuesday, September 21, 2010 1:53 AM
> To: Stevens, Weston J; ganglia-general@lists.sourceforge.net
> Subject: Re: [Ganglia-general] Does Ganglia measure itself?
>
> Hi Weston,
>
> "gmond" just looks at the low-level counters provided by the OS and has no
> awareness about its own resource usage. So, it will collect cpu-usage
> including its own cycles.
>
> Does this answer your question?
>
>
> Cheers
> Martin
> ------------------------------------------------------
> Martin Knoblauch
> email: k n o b i AT knobisoft DOT de
> www: http://www.knobisoft.de
>
>
>
> ----- Original Message ----
>> From: "Stevens, Weston J" <weston.j.stev...@boeing.com>
>> To: "ganglia-general@lists.sourceforge.net"
>><ganglia-general@lists.sourceforge.net>
>> Sent: Mon, September 20, 2010 8:20:21 PM
>> Subject: [Ganglia-general] Does Ganglia measure itself?
>>
>> For instance, if gmetad and gmond are using a few percent of CPU,
>>would this show up on the CPU usage graph? Or does it ignore itself and
>>only count everything else? Thanks
>>
>>
>> ----------------------------------------------------------------------
>> -------- Start uncovering the many advantages of virtual appliances
>> and start using them to simplify application deployment and
>> accelerate your shift to cloud computing.
>> http://p.sf.net/sfu/novell-sfdev2dev
>> _______________________________________________
>> Ganglia-general mailing list
>> Ganglia-general@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/ganglia-general
>> mG
>
> ------------------------------------------------------------------------------
> Start uncovering the many advantages of virtual appliances
> and start using them to simplify application deployment and
> accelerate your shift to cloud computing.
> http://p.sf.net/sfu/novell-sfdev2dev
> _______________________________________________
> Ganglia-general mailing list
> Ganglia-general@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ganglia-general
>
--
Jesse Becker
------------------------------------------------------------------------------
Start uncovering the many advantages of virtual appliances
and start using them to simplify application deployment and
accelerate your shift to cloud computing.
http://p.sf.net/sfu/novell-sfdev2dev
_______________________________________________
Ganglia-general mailing list
Ganglia-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ganglia-general
