Obsolete:2015 analytics datastore evaluation

This page will help us decide on a centralized data store to handle analytics at WMF.

Vision Document

https://docs.google.com/a/wikimedia.org/presentation/d/1YvswGSk7JWPbfshf3YVTLnAbG1_RQikHLZZ8sxmlX1Q/edit#slide=id.g75f4a0847_0_0

Candidates

• Spark SQL
  • this is not a datastore? Hm, I guess Hive isn't really eitiher?
• Hive on Spark
  • Not yet released.
• Hive on Tez
• OpenTSDB (HBase)
• Impala
• Cassandra
• Druid
• Apache Drill
• Presto
  • This is like Impala, but Impala would probably be easier (and maybe better)
• Mondrian + PostgreSQL
• Crate (Elastic Search)
• InfluxDB
• RESTBase
• Kudu

This page is helpful: http://blog.matthewrathbone.com/2014/06/08/sql-engines-for-hadoop.html