Use the job file and the bin script in runtime/deploy in 1.3 and 1.4-dev. Set the appropriate HADOOP_HOME var and all is good. If you don't it will complain about the env var anyway. You can execute it one any node although i tend to use the namenode for it.
On Wednesday 10 August 2011 13:36:42 jeffersonzhou wrote: > Markus, could you please be more specific? > > Do you have some links or materials? > > -----Original Message----- > From: Markus Jelsma [mailto:[email protected]] > Sent: Wednesday, August 10, 2011 5:47 PM > To: [email protected] > Subject: Re: Nutch & Hadoop > > Yes, you can simply load a job file and run it on an existing cluster. > > On Wednesday 10 August 2011 11:39:36 jeffersonzhou wrote: > > Hi, > > > > > > > > I know Nutch itself has included Hadoop. My question is if I can put > > Nutch in an Hadoop framework; that being said, Hadoop sits outside > > Nutch, and Nutch calls mapper/reducer from outside. > > > > > > > > Thanks! -- Markus Jelsma - CTO - Openindex http://www.linkedin.com/in/markus17 050-8536620 / 06-50258350
