Yes Noam. You can follow the above link to publish the nimbus, supervisor
and worker *JVM* metrics on Ganglia.

Regards,
Ankit Jain
Learning Storm
<https://www.packtpub.com/big-data-and-business-intelligence/learning-storm>


On Sun, Sep 21, 2014 at 1:16 PM, Noam Cohen <[email protected]>
wrote:

>  Thanks Ankit! That's exactly what I needed. Googling  "Storm JMXTrans
> Ganglia" got me to this page:
>
> http://jayatiatblogs.blogspot.co.il/2013/05/storm-monitoring-using-jmx-jmxtrans.html
>
> I'll try it on our cluster and let you know if it worked for me...
>
> Noam
>
>
> On 19/09/14 14:54, Ankit Jain wrote:
>
>
> Hi Noam,
>
> Yes we can easily integrate Storm with Ganglia by using the JMXTrans
> between Storm and Ganglia.
> It is also possible to publish the Storm worker metrics on Ganglia.
>
>  Regards,
> Ankit Jain
>  Learning Storm
> <https://www.packtpub.com/big-data-and-business-intelligence/learning-storm>
>
>
>
>
> On Wed, Sep 17, 2014 at 1:42 PM, Noam Cohen <[email protected]>
> wrote:
>
>> Hey guys -
>> We've just installed Storm 0.91 next to our Hadoop cluster. For Hadoop -
>> we use Ganglia extensively to get visual representation of aggregative
>> metrics (for example - number of bytes read in a given time, either from a
>> certain server or from the entire cluster, etc.). This really helps us
>> analyze performance issues that may occur in a certain time frame.
>>
>> I was wondering if there's a way to integrate Storm with Ganglia as well.
>> I mainly want to get topology-related statistics (for example - number of
>> Acked and Failed values for a certain executor or for the entire topology,
>> in a certain time frame)... Does Storm publish such metrics? Can they be
>> published to Ganglia as well?
>>
>> Thanks!
>> Noam
>>
>
>
>
> --
> Thanks,
> Ankit Jain
>
>
>


-- 
Thanks,
Ankit Jain

Reply via email to