Usage:
emr list clusters [--status=STATUS...] [--format=FORMAT]
emr list instances CLUSTERID [--status=STATUS...] [--type=TYPE...] [--format=FORMAT]
emr list steps CLUSTERID [--state=STATE...] [--format=FORMAT]
emr describe CLUSTERID [--format=FORMAT]
emr start NAME [--master=MASTER] [--node=NODE] [--count=COUNT]
emr stop CLUSTERID
emr upload FILE BUCKET BUCKETNAME
emr copy CLUSTERID BUCKET BUCKETNAME
emr run CLUSTERID BUCKET BUCKETNAME
This command is used to interface with Amazon Web Services
Elastic Map Reduce (EMR) service to run Apache Spark jobs.
It can start, list, and stop clusters and submit jobs to them.
Arguments:
CLUSTERID The AWS Cluster ID.
NAME The name of the cluster.
FILE The local file to upload.
BUCKET The name of the S3 bucket to use.
BUCKETNAME The name to file in the Bucket to use.
Options:
--status=STATUS The status to search for. [default: all]
--type=TYPE The type of instance to search for. [default: all]
--format=FORMAT How to format the output. [default: table]
--master=MASTER The type of server to use for the master node. [default: m3.xlarge]
--node=NODE The type of server to use for the worker nodes. [default: m3.xlarge]
--count=COUNT The number of servers to use [default: 3]
--state=STATE The state of the job step to filter for.
Description:
emr list clusters [--status=STATUS] [--format=FORMAT]
Lists all clusters viewable to the credentials with a
given status [default: all]. Valid statuses are: start,
boot, run, wait, terminating, shutdown, and error.
emr list instances [--status=STATUS...] [--format=FORMAT]
Lists all instances viewable to the credentials with a given
status [default: all]. Valid statuses are: start, provision,
boot, run, down. Valid types are: master, core, and task.
emr list steps CLUSTERID [--state=STATE...]
Lists all steps being performed by a cluster. Valid states are
pending, canceling, running, completed cancelled, failed, and
interrupted
emr describe CLUSTERID
Describes a cluster. Lists its status, region, type, etc.
emr stop CLUSTERID
Stops a cluster. Once a shutdown is initiated, it cannot be
undone.
emr start NAME [--master=MASTER] [--node=NODE] [--count=COUNT]
Starts a cluster with a given name, number of servers, and
server type. Bootstraps with Hadoop and Spark.
emr copy BUCKET BUCKETNAME
Copy a file from S3 to the cluster's master node.
emr run CLUSTERID BUCKET BUCKETNAME
Submit a spark application stored in an S3 bucket to the spark
cluster.