There are different ways to install Crossdata depending of your necessities. One way is using maven:
> mvn clean package -Ppackage
and it could be packaged with hive dependencies too:
> mvn clean package -Ppackage -Phive
You can build a Spark Distribution with Crossdata libraries running the make-distribution-crossdata script:
> cd scripts > ./make-distribution-crossdata.sh
- This will build Spark with the following options:
- Spark-1.5.1/hadoop2.6
- Crossdata shell
- And all available Crossdata providers
For others options run ./make-distribution-crossdata.sh --help
To run Crossdata in a standalone mode, it's necessary to install Crossdata using the make-distribution-crossdata script first. Once Crossdata is installed, just run this:
> bin/stratio-xd-shell --cassandra --mongodb
This example register the existent table "students" of the Cassandra keyspace "highschool" that has a few rows inserted. Once the table is registered then execute a query by the native way:
>xdContext.sql("CREATE TEMPORARY TABLE students USING com.stratio.crossdata.sql.sources.cassandra OPTIONS (keyspace 'highschool', table 'students', cluster 'students', pushdown 'true', spark_cassandra_connection_host '127.0.0.1')") >xdContext.sql("SELECT * FROM students").collect()
Crossdata has a Scala/Java API driver to allow to make queries programmatically on your own projects. Before do it, please see the Configuration document
Run Crossdata Server using maven:
> mvn exec:java -pl crossdata-server -Dexec.mainClass="com.stratio.crossdata.server.CrossdataApplication"
Or it is possible to start using the server jar generated previously and server configuration file:
> java -cp crossdata-server-<version>.jar com.stratio.crossdata.server.CrossdataApplication -Dcrossdata-server.external.config.filename=[path]/server-application.conf
Now that Crossdata server is running you can use the Crossdata driver importing the jar in your own project.