Using pyspark Shell¶

By default pyspark shell does not support SnappySession. To use SnappySession with pyspark shell, the shell must be started with in-memory catalog implementation.

Command to start pyspark shell with SnappySession:

./bin/pyspark --conf spark.sql.catalogImplementation=in-memory --conf spark.snappydata.connection=<locatorhost>:<clientPort>

For example:

./bin/pyspark --conf spark.sql.catalogImplementation=in-memory --conf spark.snappydata.connection=localhost:1527