Hi, Sounds like a project for Flume: http://incubator.apache.org/flume/ I don't know the if the HBase Sink/JMS Source is committed yet but this sounds like an ideal use case for them.
Disclaimer, I am a committer on Flume. Brock On Mon, May 14, 2012 at 7:17 AM, Faruk Berksöz <fberk...@gmail.com> wrote: > Dear All, > > our hadoop cluster has 1 nn , 3 dd.Evey nodes has 1 TB hard disk and 4 GB > memory. > > Hadoop version is CDH3. > > We are trying to load some records (one record has approx. 250 bytes) from > many jms queues into a hbase table. > > Is It possible to load records (For example 1.000.000 records in a second) > from many JMS queues into a HBase Table. > > What is the fastest way to do that ? > > > > Regards > > Faruk, -- Apache MRUnit - Unit testing MapReduce - http://incubator.apache.org/mrunit/