You are viewing a plain text version of this content. The canonical link for it is here.
Posted to mapreduce-user@hadoop.apache.org by Sulabh Choudhury <su...@gmail.com> on 2011/08/24 00:51:31 UTC
Capacity scheduler uses all slots from the same tasktracker
Hi,
So I just started using capacity scheduler for M/R jobs. I have 4 task
trackers each with 4 map/reduce slots.
Configured a queue so that it uses 25% (4 slots) of the available slots. I
was expecting that it would distribute the job and use slots from each of
the 4 tasktrackers but it actually uses all 4 slots from a single TT.
Is there a configuration I am missing or this is the expected behavior ?
Re: Capacity scheduler uses all slots from the same tasktracker
Posted by sulabh choudhury <su...@gmail.com>.
Well I have just 1 rack which has all the datanodes, so yes rack local.
Though the data is scattered across all the datanodes, so i guess not
"data-local".
Is there a way to force it to use slots from all the tasktrackers?
Or if I define 4 queues each of which uses 1 slot, but even then it is not
guarantedd it would use slots from different TTs
On Tue, Aug 23, 2011 at 4:07 PM, Arun C Murthy <ac...@hortonworks.com> wrote:
> Were they all 'data-local' or 'rack-local' tasks? If so, it's expected.
>
> Arun
>
>
> On Aug 23, 2011, at 3:51 PM, Sulabh Choudhury wrote:
>
> Hi,
>
> So I just started using capacity scheduler for M/R jobs. I have 4 task
> trackers each with 4 map/reduce slots.
> Configured a queue so that it uses 25% (4 slots) of the available slots. I
> was expecting that it would distribute the job and use slots from each of
> the 4 tasktrackers but it actually uses all 4 slots from a single TT.
> Is there a configuration I am missing or this is the expected behavior ?
>
>
>
>
--
--
Thanks and Regards,
Sulabh Choudhury
Re: Capacity scheduler uses all slots from the same tasktracker
Posted by Arun C Murthy <ac...@hortonworks.com>.
Were they all 'data-local' or 'rack-local' tasks? If so, it's expected.
Arun
On Aug 23, 2011, at 3:51 PM, Sulabh Choudhury wrote:
> Hi,
>
> So I just started using capacity scheduler for M/R jobs. I have 4 task trackers each with 4 map/reduce slots.
> Configured a queue so that it uses 25% (4 slots) of the available slots. I was expecting that it would distribute the job and use slots from each of the 4 tasktrackers but it actually uses all 4 slots from a single TT.
> Is there a configuration I am missing or this is the expected behavior ?
>
>