I'm trying to control the amount of memory used by Spark driver and executors -- in a normal setting, I would specify these values via the spark.driver.memory and spark.executor.memory configuration options. however, setting these in the interpreter properties in zeppelin doesn't seem to do anything -- because zeppelin also sets its own extra java flags, I couldn't simply set the -Xmx flag in spark.driver.extraJavaOptions -- the only alternative I could think of is to set the _JAVA_OPTIONS environment variable, which did the trick, but is not the optimal way I would want to use for changing the memory settings. What am I missing here? How should this be done? Thanks,
Rok