AnsweredAssumed Answered

weird failure on Map

Question asked by gabi_k on Mar 21, 2013
I can see i have 5 failures on specific Job, when i am trying to check the failure reason i have: 

    Task attempt_201302042212_11497_m_000006_0 failed to report status for 605 seconds. Killing!. Diagnostic information will be saved in /opt/mapr/hadoop/hadoop-0.20.2/bin/../logs/userlogs/job_201302042212_11497/attempt_201302042212_11497_m_000006_0/diagnostics/

all the failures on same machine (it is MapR-01 which doing CLDB and MapRed - it is small 5 nodes cluster)

when i am looking at the diagnostic dir i have 6 files:

    31151.jstack.1 , 31151.jstack.2 , 31151.jstack.3 which empty, and:

    31151.mfsthreads.1 => No Running threads
    31151.mfsthreads.2 => No Running threads
    31151.mfsthreads.3 => CONTAIN the following :

    Thread:CreateProc workarea:1c01e10 cid_working_on:2062 line:917 additional_info:, Replication info:
    stack:  fs/server/mapserver/create.cc:917
            fs/server/trans/trans.cc:1687
    Thread:ReadFile workarea:228a0c8 line:486
    stack:  fs/server/mapserver/readv3.cc:486
    Thread:ReadFile workarea:54389b8 line:486
    stack:  fs/server/mapserver/readv3.cc:486


Has someone has any idea?
Thanks.

Outcomes