Re: Hadoop internals consultant wanted

2013-05-24 Thread Stephen Boesch
Please post to a jobs forum rather than the general hadoop list. thanks. 2013/5/24 John Lilley > We are developing a YARN ApplicationMaster-based Hadoop application, and > need a consultant experienced in Hadoop internals development and YARN > specifically to help us bootstrap. Specific need

Re: MapReduce and MPI

2011-12-01 Thread Stephen Boesch
+1, I was looking at using them together just in this past week. but no bandwidth to work on myself. 2011/12/1 Mahadev Konar > Ralph, > I think there is a lot of interest in MPI over Hadoop. If you are > counting votes here's my +1. > > mahadev > > On Thu, Dec 1, 2011 at 3:31 AM, Ralph Castai

Re: [VOTE] Merge MR-279 to trunk.

2011-08-16 Thread Stephen Boesch
+1 2011/8/16 Devaraj Das > +1 (finally!) > > On Aug 16, 2011, at 2:14 PM, Mahadev Konar wrote: > > > Hi all, > > > > We are excited to let you know that we have MR-279 ready to be merged to > trunk. I have uploaded necessary details on > https://issues.apache.org/jira/browse/MAPREDUCE-279. > >

Re: Unable to build MR-279 due to unresolved ivy/maven dependencies ?

2011-04-15 Thread Stephen Boesch
I had already looked for a README (or anything similar) and it does not exist in the root dir of MR-279. Which README are you referring to? There is a README in the Yarn dir, not sure if referring to that.. but in any case attempted to follow the directions in that one. Unfortunately the mvn inst

Unable to build MR-279 due to unresolved ivy/maven dependencies ?

2011-04-14 Thread Stephen Boesch
Just checked out branch MR-279. running ant -verbose am ending up with unresolved ivy dependencies for yarn-server-common, hadoop-mapreduce-client-core and yarn-common. [ivy:resolve] :: org.apache.hadoop#yarn-server-common;1.0-SNAPSHOT: not found [ivy:resolve] :: org.apac

Re: VOTE: Committing HADOOP-6949 to 0.22 branch

2011-03-29 Thread Stephen Boesch
+1 2011/3/29 Doug Cutting > +1 > > I don't think this creates an incompatibility. It changes the RPC wire > format, but we already require that clients and servers run identical > builds. No application that ran with a prior version of Hadoop would be > broken by this change when it upgrades t