At the end of last year my colleague Steve Wooledge discussed
options you have at your disposal for querying both schema-based or self-describing structured datasources with the MapR Big Data platform. Around that time I also reviewed Open Source SQL-in-Hadoop Solutions
over at InfoQ. Today I want to summarize MapR's stance on SQL-in-Hadoop and it's one I'm very happy to share with you: you can have your cake and eat it too.
You're a heavy Hive user and want to benefit from Apache Stinger/Tez enhancements? Go for it. MapR is, at time of writing, the only Hadoop provider that ships Stinger Phase 2
on Hadoop 1.x and we will continue to deliver the latest versions. You want to use Impala? You're welcome to. On MapR, Impala
runs even more reliably and faster. Do you think Apache Drill
is the way to go? (I'm biased because I'm a Drill contributor so I won't comment on this but I think that's a great choice.) MapR is the driving force behind the creation of Drill and has several dedicated Drill engineers contributing to it along with committers from other IT companies. Likewise, if you fancy Facebook's Presto
, it also runs better on MapR. With MapR, you are in charge. You decide what you want to use to query your data with SQL; we get out of your way and focus on what we know best: providing a reliable, scalable and affordable platform with awesome support.