I've finally sorted out the display and eliminated unneeded metrics.

Have a quick read of gweb2/cluster_view.php and gweb2/host_view.php.
You will see that the "optional" graphs are loaded from
CONF_DIR/default.json
Then overlaid by CONF_DIR/cluster_<cluster-name>.json and finally
(for the host view) CONF_DIR/host_<hostname>.json.

The first step is to find the correct  CONF_DIR, it may well not be
where you expect!   Look in gweb2/conf_default.php and locate the lines

$conf['gmetad_root'] = "/var/lib/ganglia";
...
$conf['conf_dir'] = $conf['gmetad_root'] . '/conf';

So now go to /var/lib/ganglia/conf, though of course YMMV.  Have a
look at default.json which specifies the graphs to include.  In my case
I
copied it to cluster_Switches.json which will affect the cluster and
host
views in the Switches cluster.  Change included_reports to
excluded_reports and list those reports you do not want to see:

"excluded_reports":["mem_report","cpu_report"]

The unwanted graphs magically disappear! :-)  I then stopped sending
the dummy values that were only sent to keep the unwanted graphs
looking nice and removed the RRD files
(/var/lib/ganglia/rrds/Switches/...)
and have nice clean reports.

Regards,

Martin

>-----Original Message-----
>From: Rushton Martin
>Sent: 07 November 2013 11:30
>To: 'Adrian Sevcenco'
>Subject: RE: UC [Ganglia-general] UC gmetric question
>
>
>>-----Original Message-----
>>From: Adrian Sevcenco [mailto:adrian.sevce...@cern.ch]
>>Sent: 04 November 2013 13:25
>>To: ganglia-general@lists.sourceforge.net
>>Subject: Re: [Ganglia-general] UC gmetric question
>>
>>On 10/16/2013 04:00 PM, Rushton Martin wrote:
>>> You can however write a different configuration file and tell
gmetric
>>> to use it through the -c switch.  If you do this, then you will also
>>> need to start an additional daemon using -c and -p to specify the
>>> configuration file and pid-file respectively.  You can also spoof
the
>>> name and IP of the sender by using -S <IP>:<name>  I do the latter
to
>>> monitor switches.  The server gathers the metrics and sends them to
>>> gmetric using the alternative configuration file and the IP-and-name
>>> of the switch.  The alternative gmond forwards them and they appear
>>> to be a
>>So, you would have an separate cluster where it will show the default
>>metrics of the host + the switches metrics?
>
>Yes.  I have clusters named "headnodes", "compute", "switches" and
>"standalone" (for non-clustered machines).
>
>>I tried to do this but i could not eliminate the default metrics (in
>>order to not have memory and load and some other things for
>>ups/raid/switch only
>>clusters)
>
>I had a go a couple of years ago, and in the end just send standard
metrics to
>keep the old display happy.  You can be a little creative:  load_one is
(total of
>all I/O)/(theoretical maximum throughput) and proc_run is the number of
>interfaces with a non-zero output rate.  Others such as cpu_num I just
set to
>zero and ignore.
>
>I really need to revise it and write a decent front end.  When though
is
>anyone's guess.
>
>>How did you do it?
>
>I forgot to mention, I have a cron job that grabs the data (typically
via snmp),
>formats it (along with the dummy data) and sends it off.
>
>>Thanks a lot for info!
>>Adrian
>>
>>
>>
>>> separate cluster.  I very much doubt that this is what you want for
>>> PBS statistics, so leave spoofing and alternate configuration files
well alone.
>>>
>>>
>>>
>>> HTH, Martin
>>>
>>>
>>>
>>> *From:*Rita [mailto:rmorgan...@gmail.com]
>>> *Sent:* 16 October 2013 13:15
>>> *To:* ganglia-general@lists.sourceforge.net
>>> *Subject:* [Ganglia-general] gmetric question
>>>
>>>
>>>
>>> I want to implement a gmetric. To implement one, do I need to create
>>> a virtual host or something else?
>>>
>>>
>>>
>>> I am taking a look at this for example,
>>>
>>>
https://github.com/ganglia/gmetric/blob/master/hpc/pbs_jobs/pbs_jobs.
>>> s
>>> h
>>>
>>>
>>>
>>>
>>>
>>> But, what graph will this go under?  There is no specification just
a
>>> command.
>>>
>>>
>>>
>>> $GMETRIC--name PBSJOBS --type uint16 --units jobs--value $QSTAT
>>>
>>>
>>>
>>> This email and any attachments to it may be confidential and are
>>> intended solely for the use of the individual to whom it is
addressed.
>>> If you are not the intended recipient of this email, you must
neither
>>> take any action based upon its contents, nor copy or show it to
anyone.
>>> Please contact the sender if you believe you have received this
email
>>> in error. QinetiQ may monitor email traffic data and also the
content
>>> of email for the purposes of security. QinetiQ Limited (Registered
in
>>> England & Wales: Company Number: 3796233) Registered office: Cody
>>> Technology Park, Ively Road, Farnborough, Hampshire, GU14 0LX
>>> http://www.qinetiq.com
>>> <http://www.qinetiq.com/><http://www.qinetiq.com/>
>>>
>>>
>>>
>>>
---------------------------------------------------------------------
>>> -
>>> -------- October Webinars: Code for Performance Free Intel webinars
>>> can help you accelerate application performance.
>>> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the
>>> most from the latest Intel processors and coprocessors. See
abstracts
>>> and register >
>>>
http://pubads.g.doubleclick.net/gampad/clk?id=60135031&iu=/4140/ostg.
>>> c
>>> lktrk
>>>
>>>
>>>
>>> _______________________________________________
>>> Ganglia-general mailing list
>>> Ganglia-general@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/ganglia-general
>>>
>>
>>
>>--
>>----------------------------------------------
>>Adrian Sevcenco, Ph.D.                       |
>>Institute of Space Science - ISS, Romania    |
>>adrian.sevcenco at {cern.ch,spacescience.ro} |
>>----------------------------------------------

The QinetiQ e-mail privacy policy and company information is detailed elsewhere 
in the body of this email.

------------------------------------------------------------------------------
DreamFactory - Open Source REST & JSON Services for HTML5 & Native Apps
OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External API Access
Free app hosting. Or install the open source package on any LAMP server.
Sign up and see examples for AngularJS, jQuery, Sencha Touch and Native!
http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk
_______________________________________________
Ganglia-general mailing list
Ganglia-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ganglia-general

Reply via email to