Possible use case

75 views
Skip to first unread message

Gary Malouf

unread,
Mar 13, 2013, 12:34:25 PM3/13/13
to elephan...@googlegroups.com
Today, we use HBase to store user profile data.  When new data comes in, we have our map/reduce jobs read the current data from HBase in order to create 'merged' profiles when user data coming in is for users that we already know about.

If we wanted to replace HBase with ElephantDB, would the recommended process be to read from ElephantDB from within our map/reduce jobs in the same way we deal with HBase?

Jeroen van Dijk

unread,
Apr 10, 2013, 3:29:26 AM4/10/13
to elephan...@googlegroups.com

If we wanted to replace HBase with ElephantDB, would the recommended process be to read from ElephantDB from within our map/reduce jobs in the same way we deal with HBase?

To use ElephantDB data in Hadoop you use the Elephantdb shards you have on hdfs. There should be no interaction with the ElephantDB server itself. AFAIK, the easiest way is to do this through Elephantdb-cascading. I haven't used HBase so I don't know if that is different or not.

Thanks,
Jeroen
Reply all
Reply to author
Forward
0 new messages