AnsweredAssumed Answered

HBase on MapR's File System

Question asked by rthimmana on May 28, 2012
Latest reply on May 31, 2012 by Ted Dunning
Hi

We need a Distributed, Sparse, Column-Oriented Database that is very close to Google's Big Data implementation for our OLTP needs. Need to support tens of 1000s of reads, writes with ACID properties. Also, need the ability to scale up the cluster easily

Using HBase on MapR's FileSystem (as opposed to HDFS) along with MapR's Mirroring features, do we have a solution that guarantees data safety for heavy OLTP loads? Eventual consistency does not work for us. We are building a financial application where ACID properties (only one object per transaction is affected), High Availability and Consistency are MUST have requirements

Or are there any areas where further work needs to be done by MapR? If HBase and OLTP is a bad idea are there any other options out there - commercial and open source - both?

Regards
Ramesh

Outcomes