I've been having a problem for the past few weeks. I will kick off a
job that will have a bunch of map tasks. At some point in the job
(sometimes at 1%, sometimes at 40%) the mappers will start reporting
the following error "Task attempt_201009021455_0033_m_000000_0 failed
to report status for 600 seconds. Killing!" The mapper progress will
sit at 100%. Eventually, too many mappers fail, and the job fails.

I have added context.progress(); to the map method, based on some
reading that I've done, but it doesn't help.

I would appreciate any help you can provide me.

Thank you!

Does your mapper access external resources which may take some time to
return ?

On Thu, Sep 9, 2010 at 11:28 AM, Pavel Gutin <[EMAIL PROTECTED]> wrote:

> I've been having a problem for the past few weeks. I will kick off a
> job that will have a bunch of map tasks. At some point in the job
> (sometimes at 1%, sometimes at 40%) the mappers will start reporting
> the following error "Task attempt_201009021455_0033_m_000000_0 failed
> to report status for 600 seconds. Killing!" The mapper progress will
> sit at 100%. Eventually, too many mappers fail, and the job fails.
>
> I have added context.progress(); to the map method, based on some
> reading that I've done, but it doesn't help.
>
> I would appreciate any help you can provide me.
>
> Thank you!
>