Hello Lewis, trunk runs fine on a Hadoop 2.6 cluster. We have not seen any issues. I did some attempts at first to port some jobs to the new mapreduce API, but there were problems and some API's didn't exist in the new mapreduce API. As far as i know the old mapred API is not going to be deprecated for a long while to come.
Re: Nutch 3.0, i see no point in that so far. Markus -----Original message----- From: Lewis John Mcgibbney<[email protected]> Sent: Thursday 25th June 2015 1:19 To: [email protected] Subject: [IMPORTANT] Migration Towards HAdoop 2.X --> 3.X Hi Folks, In not too long time Hadoop will be up at 3.X for stable official releases. I wanted to solicit the dev@ community to see what difficulties if any people have had running Nutch trunk on Hadoop 2.X. Hadoop 2.X is supported on Nutch 2.X but getting the patches all correct is literally a PITA... we are working on that down in the Gora community and need to get a better more frequent release cycle. I just wanted to know if there was motivation for us to get some patches committed to trunk, releases it as 1.11 then focus the next development drive on a switch to Hadoop 2.X for trunk. We could potentially then release Nutch > 1.11 as 3.0. What do you guys think? Thanks Lewis -- Lewis
