Configuration

---

BigDL uses Java properties to control its behavior. Here's the list of these properties.

How to set the properties

Spark

If you run BigDL on Apache Spark, you can set the properties by passing spark-submit options. Here's an example:

# Say you want to set property FOO to value BAR

spark-submit ...
    --conf 'spark.executor.extraJavaOptions=-DFOO=BAR' # Set that property for executor process
    --conf 'spark.driver.extraJavaOptions=-DFOO=BAR'   # Set that property for driver process
    ...

Local Java/Scala program

If you run BigDL in a local Java/Scala program, you can set the properties by passing JVM parameters. Here's an example:

# Say you want to set property FOO to value BAR
java -cp xxx.jar -DFOO=BAR your.main.class.name

Available Properties

Category	Property	Default value	Description
Logging	bigdl.utils.LoggerFilter.disable	false	Disable redirecting logs of Spark and BigDL to a file.
	bigdl.utils.LoggerFilter.logFile	Current_Working_Directory/bigdl.log	Where is the redirecting log.
	bigdl.utils.LoggerFilter.enableSparkLog	true	Enable redirecting Spark logs to logFile. Set it to false when you don't want to see Spark logs in the redirecting log file.
Mode	bigdl.localMode	false	Whether BigDL is running as a local Java/Scala program.
Multi-threading	bigdl.coreNumber	half of the virtual core number	How many cores BigDL use on your machine. It is only used when bigdl.localMode is set to true. If hyper thread is enabled on your machine, DO NOT set it larger than half of the virtual core number.
	bigdl.Parameter.syncPoolSize	4	Thread pool size for syncing parameter between executors.
Distributed Training	bigdl.network.nio	true	Whether use NIO as BlockManager backend in Spark 1.5. If it is set to false, user can specify spark.shuffle.blockTransferService to change the BlockManager backend. ONLY used when running on Spark 1.5.
	bigdl.failure.retryTimes	5	How many times to retry when there's failure in distributed Training.
	bigdl.failure.retryTimeInterval	120	Unit is second. How long recount the retry times.
	bigdl.check.singleton	false	Whether to check if multiple partition run on a same executor, which is bad for performance.
Tensor	bigdl.tensor.fold	1000	How many elements in a tensor to determine it is a large tensor, and thus print only part of it.