Hi, Correct me if I'm wrong but you should very much be able to get the currently running jobs via the provided API. If you call JobTracker.getAllJobs() you'll get an array of JobStatus objects, which each correspond to a job. You can then get the job progress via JobStatus.mapProgress and JobStatus.reduceProgress.
The one caveat is that you can't link into a running JobTracker, so you only have access to it if you launch it from your own code. However! You can (and should) use JobClient ( http://hadoop.apache.org/common/docs/r0.19.1/api/org/apache/hadoop/mapred/JobClient.html) instead to query information about jobs on your cluster. You can instantiate these objects with some basic configuration information, and then you will have access to JobClient.getAllJobs much like the aforementioned JobTracker.getAllJobs. -Toby On Mon, Nov 9, 2009 at 12:15 PM, Todd Lipcon <t...@cloudera.com> wrote: > Hi, > > No, there is currently no public-facing API for tracking job status aside > from parsing those JSPs (or adding your own) > > -Todd > > On Mon, Nov 9, 2009 at 5:28 AM, pnd <prafulla.daw...@gmail.com> wrote: > > > > > Hi i am new to hadoop been using it for few weeks now, i tried few > > map/reduce example and could see logs in jobtracker.jsp. > > Is there any way to use jobtracker > > api( > > > http://hadoop.apache.org/common/docs/r0.19.1/api/org/apache/hadoop/mapred/JobTracker.html > > ) > > for tracking the jobs instead using jobtracker jsp because I am trying to > > build new UI to show the job progress according to my requirements. > > > > Thanks in advance. > > > > -- > > View this message in context: > > > http://old.nabble.com/Using-jobtracker-api-for-developing-UI-in-hadoop-0.19.1-tp26266407p26266407.html > > Sent from the Hadoop core-user mailing list archive at Nabble.com. > > > > >
