TonY Configurations

Application Properties:

Name	Default	Meaning
tony.other.namenodes		Namenode URIs to get delegation tokens from.
tony.yarn.queue	default	Default queue to submit to YARN.
tony.application.name	TensorFlowApplication	Name of your YARN application.
tony.application.node-label		YARN partition which this application should run in.
tony.application.single-node	false	Whether this is single node training or not.
tony.application.enable-preprocess	false	Whether the AM should invoke the user's python script or not.
tony.application.timeout	0	Max runtime of the application before killing it, in milliseconds.

Task Properties:

Name	Default	Meaning
tony.task.executor.jvm.opts	-Xmx1536m	JVM opts for each TaskExecutor.
tony.task.registration-timeout-sec	300	Timeout, in seconds, for AM to resubmit unregistered tasks (or fail if no retries configured).
tony.task.registration-retry-count	0	How many times we should resubmit unregistered tasks after the timeout interval.
tony.task.heartbeat-interval	1000	Frequency, in milliseconds, for which TaskExecutors should heartbeat with AM.
tony.task.max-missed-heartbeats	25	How many missed heartbeats before declaring a TaskExecutor dead.

AM Configuration

Name	Default	Meaning
tony.am.retry-count	0	How many times a failed AM should retry.
tony.am.memory	2g	AM memory size, requested as a string (e.g. '2g' or '2048m').
tony.am.vcores	1	Number of AM vcores to use.
tony.am.gpus	0	Number of AM GPUs to use. (In general, should only be applicable in single node mode.)

Executor Configuration

Name	Default	Meaning
tony.ps.memory	2g	Parameter server memory size, requested as a string (e.g. '2g' or '2048m').
tony.ps.vcores	1	Number of vcores per parameter server.
tony.ps.instances	1	Number of parameter servers to request.
tony.ps.instances	0	Timeout, in milliseconds for the user's python processes before forcibly killing them.
tony.worker.memory	2g	Worker memory size, requested as a string (e.g. '2g' or '2048m').
tony.worker.vcores	1	Number of vcores per worker.
tony.worker.gpus	0	Number of GPUs per worker.
tony.worker.instances	1	Number of workers to request.

Others

Name	Default	Meaning
tony.application.security.enabled	true	Whether this application is running in a Kerberized grid. Setting this to true will fetch tokens from the cluster as well as between the client and AM.
tony.application.hdfs-conf-path		Path to HDFS configuration, to be passed as an environment variable to the python training scripts.
tony.application.yarn-conf-path		Path to YARN configuration, to be passed as an environment variable to the python training scripts.