Thanks Mohammed, I was aware of Calliope, but haven't used it since with since the spark-cassandra-connector project got released. I was not aware of the CalliopeServer2; cool thanks for sharing that one.
I would appreciate it if you could lmk how you decide to proceed with this; I can see this coming up on my radar in the next few months; thanks. -Todd On Fri, Apr 3, 2015 at 5:53 PM, Mohammed Guller <moham...@glassbeam.com> wrote: > Thanks, Todd. > > > > It is an interesting idea; worth trying. > > > > I think the cash project is old. The tuplejump guy has created another > project called CalliopeServer2, which works like a charm with BI tools that > use JDBC, but unfortunately Tableau throws an error when it connects to it. > > > > Mohammed > > > > *From:* Todd Nist [mailto:tsind...@gmail.com] > *Sent:* Friday, April 3, 2015 11:39 AM > *To:* pawan kumar > *Cc:* Mohammed Guller; user@spark.apache.org > > *Subject:* Re: Tableau + Spark SQL Thrift Server + Cassandra > > > > Hi Mohammed, > > > > Not sure if you have tried this or not. You could try using the below api > to start the thriftserver with an existing context. > > > https://github.com/apache/spark/blob/master/sql/hive-thriftserver/src/main/scala/org/apache/spark/sql/hive/thriftserver/HiveThriftServer2.scala#L42 > > The one thing that Michael Ambrust @ databrick recommended was this: > > You can start a JDBC server with an existing context. See my answer here: > http://apache-spark-user-list.1001560.n3.nabble.com/Standard-SQL-tool-access-to-SchemaRDD-td20197.html > > So something like this based on example from Cheng Lian: > > > * Server* > > import org.apache.spark.sql.hive.HiveContext > > import org.apache.spark.sql.catalyst.types._ > > > > val sparkContext = sc > > import sparkContext._ > > val sqlContext = new HiveContext(sparkContext) > > import sqlContext._ > > makeRDD((1,"hello") :: (2,"world") > ::Nil).toSchemaRDD.cache().registerTempTable("t") > > // replace the above with the C* + spark-casandra-connectore to generate > SchemaRDD and registerTempTable > > > > import org.apache.spark.sql.hive.thriftserver._ > > HiveThriftServer2.startWithContext(sqlContext) > > Then Startup > > ./bin/beeline -u jdbc:hive2://localhost:10000/default > > 0: jdbc:hive2://localhost:10000/default> select * from t; > > > > I have not tried this yet from Tableau. My understanding is that the > tempTable is only valid as long as the sqlContext is, so if one terminates > the code representing the *Server*, and then restarts the standard thrift > server, sbin/start-thriftserver ..., the table won't be available. > > > > Another possibility is to perhaps use the tuplejump cash project, > https://github.com/tuplejump/cash. > > > > HTH. > > > > -Todd > > > > On Fri, Apr 3, 2015 at 11:11 AM, pawan kumar <pkv...@gmail.com> wrote: > > Thanks mohammed. Will give it a try today. We would also need the > sparksSQL piece as we are migrating our data store from oracle to C* and it > would be easier to maintain all the reports rather recreating each one from > scratch. > > Thanks, > Pawan Venugopal. > > On Apr 3, 2015 7:59 AM, "Mohammed Guller" <moham...@glassbeam.com> wrote: > > Hi Todd, > > > > We are using Apache C* 2.1.3, not DSE. We got Tableau to work directly > with C* using the ODBC driver, but now would like to add Spark SQL to the > mix. I haven’t been able to find any documentation for how to make this > combination work. > > > > We are using the Spark-Cassandra-Connector in our applications, but > haven’t been able to figure out how to get the Spark SQL Thrift Server to > use it and connect to C*. That is the missing piece. Once we solve that > piece of the puzzle then Tableau should be able to see the tables in C*. > > > > Hi Pawan, > > Tableau + C* is pretty straight forward, especially if you are using DSE. > Create a new DSN in Tableau using the ODBC driver that comes with DSE. Once > you connect, Tableau allows to use C* keyspace as schema and column > families as tables. > > > > Mohammed > > > > *From:* pawan kumar [mailto:pkv...@gmail.com] > *Sent:* Friday, April 3, 2015 7:41 AM > *To:* Todd Nist > *Cc:* user@spark.apache.org; Mohammed Guller > *Subject:* Re: Tableau + Spark SQL Thrift Server + Cassandra > > > > Hi Todd, > > Thanks for the link. I would be interested in this solution. I am using > DSE for cassandra. Would you provide me with info on connecting with DSE > either through Tableau or zeppelin. The goal here is query cassandra > through spark sql so that I could perform joins and groupby on my queries. > Are you able to perform spark sql queries with tableau? > > Thanks, > Pawan Venugopal > > On Apr 3, 2015 5:03 AM, "Todd Nist" <tsind...@gmail.com> wrote: > > What version of Cassandra are you using? Are you using DSE or the stock > Apache Cassandra version? I have connected it with DSE, but have not > attempted it with the standard Apache Cassandra version. > > > > FWIW, > http://www.datastax.com/dev/blog/datastax-odbc-cql-connector-apache-cassandra-datastax-enterprise, > provides an ODBC driver tor accessing C* from Tableau. Granted it does not > provide all the goodness of Spark. Are you attempting to leverage the > spark-cassandra-connector for this? > > > > > > > > On Thu, Apr 2, 2015 at 10:20 PM, Mohammed Guller <moham...@glassbeam.com> > wrote: > > Hi – > > > > Is anybody using Tableau to analyze data in Cassandra through the Spark > SQL Thrift Server? > > > > Thanks! > > > > Mohammed > > > > > > >
