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 Harsh J <qw...@gmail.com> on 2011/03/15 17:39:57 UTC

Re: Frustrating "pending" statuses

[ Moving this to mapreduce-user@ - since it isn't a common-development
related question, but a mapreduce specific one || bcc-ing common-dev@
]

On Tue, Mar 15, 2011 at 1:25 PM, Keith Wiley <kw...@keithwiley.com> wrote:
> Despite plenty of available mapper slots (as inferred by "Map Task Capacity" minus "Maps" on the jobtracker) our jobs often linger with all or many of the map tasks stuck in a pending status for five to ten minutes.  Eventually they start, but why the seemingly unnecessary delay?  There also might be a correlation between starting an unrelated job to "unstick" the pending tasks of the previous job...although this correlation is difficult to quantify; it may be a coincidence.

Could we additionally know what scheduler is in effect here, and does
the log of JobTracker contain anything you think may be remotely
related to this issue?

-- 
Harsh J
http://harshj.com