hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From llpind <sonny_h...@hotmail.com>
Subject HBase Map/Reduce
Date Mon, 01 Jun 2009 21:53:17 GMT

Hey All,

I'm new to map/reduce & HBase.   Sorry if this has been asked before.  I
would like to run a map/reduce job on a Hadoop (0.19.1)/Hbase (0.19.3)
cluster.  I have attached the modified version of SampleUploader &
DBInputFormat. 

When I run the uploader program from my windows box (within eclipse, jars
/hbase-site.xml in build path) the map phase takes a really long time, and
the reduce fails during intensive write operations (which may be solved now
since I've added more boxes).  It's not run in a disturbuted manner (as far
as I can tell).  I'd like to move it to one of the linux boxes, and run on
the cluster.  Also it seems like it doesn't split the DB results. 

I'm hoping someone can suggest how I should either change the program, or
move where it's run so it's disturbuted.

I have 8 slaves, and 1 master.

Thanks.
http://www.nabble.com/file/p23823017/HBaseDBUploader.java
HBaseDBUploader.java 
http://www.nabble.com/file/p23823017/DBInputFormat.java DBInputFormat.java 
-- 
View this message in context: http://www.nabble.com/HBase-Map-Reduce-tp23823017p23823017.html
Sent from the HBase User mailing list archive at Nabble.com.


Mime
View raw message