We joined forces with MapR Technologies last June to deliver enterprise-grade Hadoop on EMR with their M5 and M3 Editions. Today we're making MapR's M7 Edition available on EMR, enabling users to run 24x7 HBase applications in addition to their Hadoop ones. The M7 architecture provides the following advantages for HBase users:
- Up to 100K ops/s per node on hs1 instances.
- No compactions.
- Seamless region splits.
- Instant recovery from any failure.
- Consistent low latency.
- Full HA.
- Point-in-time recovery (consistent snapshots).
- Disaster recovery (mirroring).
MapR is the only distribution that enables Linux applications and commands to access data directly in the cluster via the NFS interface that is available with all MapR editions. MapR M7 was optimized for cloud deployments including high performing instances such as High Storage and High I/O.
To launch an M7 cluster, select the MapR M7 Edition in the EMR New Job Flow Wizard:
You can also use the elasticmapreduce CLI. To launch the latest version of M7 on EMR, use the following command:
Visit the MapR M7 documentation to learn more.
PS - If you don't need the full feature set offered by M7, you can now run MapR M5 at prices that have been reduced by 13% to 45%, depending on instance size.