Review comments

docs/source/gs-processing/usage/emr.rst

@@ -110,7 +110,7 @@ The script has four required arguments:
   ``--gsp-arguments "--input-config gsp-config.json --input-prefix s3://my-bucket/raw-data [...]"``.
 * ``--instance-type``: The instance type to use for our cluster. Our script only supports
   a uniform instance types currently.
-* ``--worker-count``: Number of worker instances to launch for the cluster.
+* ``--instance-count``: Number of worker instances to launch for the cluster.
 
 Run ``python graphstorm-processing/scripts/submit_gsp_emr_step.py --help`` for more optional arguments.
 
@@ -136,7 +136,8 @@ using the above Python script.
     GENERATE_REVERSE="true"
 
 
-    # We assume this script is saved in the same path as submit_gsp_emr_step.py
+    # We assume this script is saved in the same path as submit_gsp_emr_step.py,
+    # that is graphstorm-processing/scripts
     SCRIPT_DIR=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
     # Upload the entry point to S3
     PATH_TO_ENTRYPOINT="$SCRIPT_DIR/../graphstorm_processing/distributed_executor.py"

graphstorm-processing/graphstorm_processing/constants.py

@@ -66,7 +66,7 @@ class ExecutionEnv(Enum):
     LOCAL = 1
     SAGEMAKER = 2
     EMR_SERVERLESS = 3
-    EMR_ON_EC2 = 3
+    EMR_ON_EC2 = 4
 
 
 ################# Supported filesystem types#############

graphstorm-processing/scripts/submit_gsp_emr_step.py

@@ -57,7 +57,7 @@ def parse_args():
     )
     parser.add_argument("--timeout-hours", type=int, default=1)
     parser.add_argument(
-        "--worker-count",
+        "--instance-count",
         type=int,
         required=True,
         help="Number of worker instances. Required",
@@ -127,7 +127,7 @@ def main():
     image_tag = args.gsp_image_tag
     timeout_hours = args.timeout_hours
     instance_type = args.instance_type
-    core_instance_count = args.worker_count
+    core_instance_count = args.instance_count
 
     s3_entry_point = args.entry_point_s3