Elastic High Performance Computing:SubmitJob

The cluster ID.

You can call the ListClusters operation to query the cluster ID.

The command that is used to run the job.

The name of the user that runs the job.

You can call the ListUsers operation to query the users of the cluster.

The password that corresponds to the username.

The name of the job. The name must be 6 to 30 characters in length and start with a letter. It can contain letters, digits, and periods (.).

The priority of the job. Valid values: 0 to 9. A larger value indicates a higher priority.

Default value: 0.

The path that is used to run the job.

The output file path of stdout.

The output file path of stderr.

Specifies whether the job can be rerun. Valid values:

• true: The job can be rerun.
• false: The job cannot be rerun.

The job array.

Format: X-Y:Z. The minimum index value X is the first index. The maximum index value Y is the last index. Z is the step size. For example, 2-7:2 indicates that three jobs need to be run and their index values are 2, 4, and 6.

The runtime variables passed to the job. They can be accessed by using environment variables in the executable file.

The URL of the job file that is uploaded to an Object Storage Service (OSS) bucket.

The command for file decompression. The command that is used to decompress the job files downloaded from an OSS bucket. Valid values:

• tar xzf: Decompresses GZIP files.
• tar xf: Decompresses TAR files.
• unzip: Decompresses ZIP files.

The command to perform on the job after the job is submitted.

The ID of the containerized application. If you want to use a container application, you must specify its ID.

You can call the ListContainerApps operation to query the container application ID.

The name of the queue in which the job is run.

You can call the ListQueues operation to query the name of the queue.

The number of compute nodes required to run the job.

The number of CPU cores required by a single compute node.

The number of processes created for a single compute node.

The parameter is applicable to Message Passing Interface (MPI) jobs.

The number of threads created for a single compute node.

The parameter is applicable to OpenMP jobs.

The maximum memory usage required by a single compute node. Unit: GB, MB, or KB. The unit is case-insensitive.

The maximum GPU usage required by a single compute node.

The parameter takes effect only when the cluster uses PBS and a compute node is a GPU-accelerated instance.

The maximum running time of the job. Valid formats:

• hh:mm:ss
• mm:ss
• ss

We recommend that you use the hh:mm:ss format. If the maximum running time is 12 hours, set the value to 12:00:00.

The number of retries for the job. Valid values: 1 to 10. You can only retry jobs that are run on the PBS clusters.

The priority of the job retry. Valid values: 0 to 9. A larger value indicates a higher priority.

The retry condition of the job. If the exit code is the value of the parameter, the job retry is triggered.

Specifies whether to use an asynchronous link to submit the job.

Default value: false.