By "ThriftServer," you mean HiveServer2? Or are you talking about HBase's thrift gateway? Both should work, though neither are required. As my previous mail, you should have better scalable performance by using HBase's InputFormats instead of these intermediary services.
I have no experience running HS2 or the thrift gateway with YARN or Mesos. Doesn't mean it's not possible... On Tuesday, February 17, 2015, Marco <[email protected]> wrote: > In the meantime I've played with Spark 'standalone' cluster. There it > works. > > Is my assumption correct that the Thriftserver only works with the Spark > standalone cluster ? Or would Mesos be an alternative? > > BR Marco > > 2015-02-16 16:36 GMT+01:00 Marco <[email protected] <javascript:;>>: > > > Hi, > > > > I've played with the feature to expose RDD via Thrift to enable JDBC > > access. (Spark 1.2) > > > > > > val eventsView = sqlContext.createSchemaRDD(eventSchemaRdd) > > eventsView.registerTempTable("Events") > > > > HiveThriftServer2.startWithContext(sqlContext) > > > > > > This works all fine. > > > > Now, my understanding is you can't deploy this to a yarn-cluster. Is this > > correct or what are my options here ? My major concern is scaleability > > (e.g. having a lot of SQL requests, which may also not be trivial) > > > > Thanks, > > Marco > > > > > > > > > -- > Viele Grüße, > Marco >
