FAQ
I'm seeing some really bizarre behavior from Hadoop 0.19.1. I have a
fairly large job with about 29000 map tasks and 72 reducers. there are
304 map task slots in the cluster. When the job starts, it runs 304
map tasks at a time. As time goes on the number of map tasks run
concurrently drops. For at least half of the execution exactly 152
mappers were run at a time. Towards, the end , when there were only
100 or so tasks remaining, the number of concurrent mappers quickly
fell to 2 at a time, bringing the end of the map phase to a crawl.
This was the only job running on the cluster. Has anyone else seen
behavior like this?

Search Discussions

  • Runping Qi at Mar 4, 2009 at 1:17 am
    Were task Trackers black-listed?

    On Tue, Mar 3, 2009 at 3:25 PM, Nathan Marz wrote:

    I'm seeing some really bizarre behavior from Hadoop 0.19.1. I have a fairly
    large job with about 29000 map tasks and 72 reducers. there are 304 map task
    slots in the cluster. When the job starts, it runs 304 map tasks at a time.
    As time goes on the number of map tasks run concurrently drops. For at least
    half of the execution exactly 152 mappers were run at a time. Towards, the
    end , when there were only 100 or so tasks remaining, the number of
    concurrent mappers quickly fell to 2 at a time, bringing the end of the map
    phase to a crawl. This was the only job running on the cluster. Has anyone
    else seen behavior like this?
  • Nathan Marz at Mar 4, 2009 at 1:49 am
    Nope... and there were no failed tasks.

    On Mar 3, 2009, at 5:16 PM, Runping Qi wrote:

    Were task Trackers black-listed?

    On Tue, Mar 3, 2009 at 3:25 PM, Nathan Marz wrote:

    I'm seeing some really bizarre behavior from Hadoop 0.19.1. I have
    a fairly
    large job with about 29000 map tasks and 72 reducers. there are 304
    map task
    slots in the cluster. When the job starts, it runs 304 map tasks at
    a time.
    As time goes on the number of map tasks run concurrently drops. For
    at least
    half of the execution exactly 152 mappers were run at a time.
    Towards, the
    end , when there were only 100 or so tasks remaining, the number of
    concurrent mappers quickly fell to 2 at a time, bringing the end of
    the map
    phase to a crawl. This was the only job running on the cluster. Has
    anyone
    else seen behavior like this?
  • Pedro Vivancos at Mar 4, 2009 at 8:35 am
    I have the same problem... Please let me know if you solve it and how to.

    Thanks.


    On Wed, Mar 4, 2009 at 2:49 AM, Nathan Marz wrote:

    Nope... and there were no failed tasks.



    On Mar 3, 2009, at 5:16 PM, Runping Qi wrote:

    Were task Trackers black-listed?

    On Tue, Mar 3, 2009 at 3:25 PM, Nathan Marz wrote:

    I'm seeing some really bizarre behavior from Hadoop 0.19.1. I have a
    fairly
    large job with about 29000 map tasks and 72 reducers. there are 304 map
    task
    slots in the cluster. When the job starts, it runs 304 map tasks at a
    time.
    As time goes on the number of map tasks run concurrently drops. For at
    least
    half of the execution exactly 152 mappers were run at a time. Towards,
    the
    end , when there were only 100 or so tasks remaining, the number of
    concurrent mappers quickly fell to 2 at a time, bringing the end of the
    map
    phase to a crawl. This was the only job running on the cluster. Has
    anyone
    else seen behavior like this?

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
groupcommon-user @
categorieshadoop
postedMar 3, '09 at 11:26p
activeMar 4, '09 at 8:35a
posts4
users3
websitehadoop.apache.org...
irc#hadoop

People

Translate

site design / logo © 2022 Grokbase