Hi,
Well, I am assuming you have one central node collecting data from all your
servers. On that server, you will run 2 gmond instances, setting each one
up with a UDP port to listen on, that uses the "udp_recv_channel" stanza,
where you can set the IP address and port to bind to. You can achieve this
by copying your current gmond.conf (presumably in /etc/ganglia/), and
running gmond -c <path to config>. You will need to change one of these
gmond config files, to bind to either a different port, or different IP,
otherwise the second one will fail to start.
Once both gmond instances are running, you define 2 data sources in your
gmetad.conf. You will already have one in there, something like
'data_source "my cluster" localhost'. You add another line like
'data_source "my other cluster" localhost <port number>', you will need to
insert the port number you configured in your second gmond.conf. If the
second gmond is running on a different IP or host, then you will need to
change localhost to point at the other IP.
Cheers, Michael.
On 16 October 2013 14:18, Song Sam <[email protected]> wrote:
> cluster is just a logical separation of hosts.as far as i known, you only
> need to configure the cluster section of gmond.conf in each host.
> 在 2013-10-16 上午8:47，"Rita" <[email protected]>写道：
>
>
>>
>>
>> On Tue, Oct 15, 2013 at 5:50 PM, Michael Shearer <[email protected]
>> > wrote:
>>
>>> Hi,
>>>
>>> You should be able to setup 2 gmond, one for each set of servers, and
>>> then specify 2 data sources in your gmetad config. Just change the listen
>>> port if you run both gmond on the same server/ IP.
>>>
>>> Cheers, Michael.
>>>
>>>
>>> On 16 October 2013 00:55, Rita <[email protected]> wrote:
>>>
>>>> Using 3.5.7 and I have 2 sets of machines I would like to monitor. Set
>>>> 1, 10 servers and set 2, 16 servers. Currently, they are all pointing to
>>>> the same gmetad. I am using Unicast therefore I would like to classify
>>>> these into two sets. Do I need to run two separate gmetads?
>>>>
>>>> I have read the following docs but haven't had much luck
>>>> http://jablonskis.org/2011/monitoring-multiple-clusters-using-ganglia/
>>>>
>>>>
>>>> --
>>>> --- Get your facts first, then you can distort them as you please.--
>>>>
>>>>
>>>> ------------------------------------------------------------------------------
>>>> 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.clktrk
>>>> _______________________________________________
>>>> Ganglia-general mailing list
>>>> [email protected]
>>>> https://lists.sourceforge.net/lists/listinfo/ganglia-general
>>>>
>>>>
>>>
>>
>> Thanks for your response. Do you happen to have an example? I have tried
>> searching on Google but couldn't find anything relevant.
>>
>>
>>
>> --
>> --- Get your facts first, then you can distort them as you please.--
>>
>>
>> ------------------------------------------------------------------------------
>> 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.clktrk
>> _______________________________________________
>> Ganglia-general mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/ganglia-general
>>
>>
>
> ------------------------------------------------------------------------------
> 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.clktrk
> _______________________________________________
> Ganglia-general mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/ganglia-general
>
>
------------------------------------------------------------------------------
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.clktrk
_______________________________________________
Ganglia-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ganglia-general
