You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@falcon.apache.org by Mahak Mukhi <mm...@yahoo-inc.com.INVALID> on 2015/06/12 00:28:54 UTC

Issues with Falcon performance testing

Hi,
I have the following scenario; 2 clusters; source(about 16 nodes)  and target(about 14 nodes), feeds set up for replication with a 10 minute frequency. Everything works fine when the number of feeds is less than 3. When the third feed is added everything just hangs up. Oozie shows the jobs to be running but they never complete. Does anyone know what might be the issue here? Has anybody else run into a situation like this? 
Also, is there a performance evaluation report for falcon that I can refer to? 
Regards,
Mahak Mukhi

Re: Issues with Falcon performance testing

Posted by Arpit Gupta <ar...@hortonworks.com>.
Did you check your RM to make sure you have enough memory to launch more AM containers as well as containers for task. If oozie is showing the jobs as running but they are not making progress my first guess would be that yarn configurations need to be tuned.


If you log on to your RM web console you will be able to see how many applications are in pending state.

--
Arpit Gupta
Hortonworks Inc.
http://hortonworks.com/

> On Jun 11, 2015, at 3:28 PM, Mahak Mukhi <mm...@yahoo-inc.com.INVALID> wrote:
> 
> Hi,
> I have the following scenario; 2 clusters; source(about 16 nodes)  and target(about 14 nodes), feeds set up for replication with a 10 minute frequency. Everything works fine when the number of feeds is less than 3. When the third feed is added everything just hangs up. Oozie shows the jobs to be running but they never complete. Does anyone know what might be the issue here? Has anybody else run into a situation like this? 
> Also, is there a performance evaluation report for falcon that I can refer to? 
> Regards,
> Mahak Mukhi