12.2.4. Setting Up a Batch Job

When you start a batch job using the Batch job application, or by executing a batch job template, the job wizard launches:

  1. To apply a tag to the job, select the tag from the Tag drop box. Tags provide a way to categorize jobs; for example, by type, owner, or billing department. They can also be used to predetermine the job working directory and the repository folder into which files will be imported after job completion.

    To create a tag:

    1. From the Tag drop box, select Create new tag.

    2. In the Tag edit box, specify the tag name. The specified name will be used to identify the tag in the tag list, and will also be automatically applied to the job name.

    3. In the Cluster job name edit box, specify the job name or account name that will populate the Cluster job name field in the job submission wizard, if such a field has been enabled by an EKM administrator. If a cluster job name is specified in the job submission wizard, the name will be reported in the RSM job log, and will provide a traceable piece of data that administrators can use to track job submissions to a cluster (for billing purposes, for example). For information on adding a cluster job name field to the job wizard, see Tracking Cluster Usage Through a Cluster Job Name).

    4. To set a specific working directory for jobs using this tag, click Browse below Working Directory, then select the desired directory on the EKM server. The working directory is where job files are written, and in most cases, where the job is run.

    5. To set the repository folder into which files will be imported after a job has executed, click Browse below Import files after job completion to, then select the desired folder.

    6. Click OK to add the tag to the Tag list.

    To view a tag's properties, or delete a tag, select Manage tags from the Tag drop box.

  2. Accept or edit the job name. If you have applied a tag to the job, the job name defaults to the tag name, but you can override this if desired.

    Note that an additional field may appear below the Job name field if the EKM administrator has configured a cluster job name field for the job submission wizard. For more information see Tracking Cluster Usage Through a Cluster Job Name in the Administration Guide.

  3. Specify the job Working Directory. This is the EKM server directory to which job files will be written, and in most cases, where the job will run. If you have applied a tag to the job, then the working directory will be automatically set to the directory associated with the tag.

    The options available depend on which view was current when you started the job:

    • Use current directory [directory name]. Available if the current view is the File System view, this sets the working directory to the current folder in the File System view. This option is convenient when you want to start a job from an existing directory that contains the desired input files.

    • Use working directory of current job [directory name]. Available if the current view is a job view, this runs the job from the working directory of the job that you are viewing. This option is convenient if you want the new job to use output files from a previous job.

    • Let EKM create working directory. EKM will automatically create a working directory in the EKM Job Data Directory (to see where this is located, refer to EKM and RSM Job Directories).

    • Choose a directory on the server. Click Browse to display the folder structure of the EKM server. When you select a folder, its path will be displayed in the edit box.


      Note:
      • If selecting the working directory of an existing job, that directory will not be deleted until the last job that is using that directory is deleted.

      • If selecting a directory on the server, make sure that you do not select a directory that could potentially be deleted. For example, do not select another user’s job working directory, as it may be deleted if the user deletes the job with which it is associated.

      • When selecting a job directory, you must have write permission on that directory.


    The working directory will be created when you advance to the next screen. A job object will also be created in the Job Monitor at this time. Note that if you were to close the dialog box now or in the next step, the job would be cancelled, but the job object and working directory would remain.

  4. Click Next.

    At this point, a job is created in the Job Monitor, and a working directory is created. Note that if you were to close the dialog box now or in the next step, the job would be cancelled, but the job object and working directory would remain.

    You can now upload input files to the working directory. If you started the job from existing repository files, the files will begin uploading to the working directory automatically.

    • To upload files from your local computer, select the Local System tab. If the files are currently displayed in a window such as Windows Explorer, or are on your desktop, you can drag them into the dialog. Otherwise, click Add and select the files in the Open dialog box. In Firefox and Chrome browsers you can select multiple files simultaneously. To select consecutive files, hold down the Shift key while selecting files. To select non-consecutive files, hold down the Ctrl key while selecting files. (In Internet Explorer you can only select a single file.)

      If you would like to use the File Transfer Client to upload files, click the Launch File Transfer Client link. See Uploading Files Using the File Transfer Client for more information.

    • To upload files from the EKM repository, select the Repository tab. Click Add to display the repository file system in a pop-up window. Navigate to the repository folder where the files reside, or enter the path to the folder in the address field and click Go. To move up a level in the folder structure, click the up arrow next to the Go button. Select the files that you want to upload by enabling their check boxes, then click OK to add the selected files to the upload list.


      Note:
      • The Repository tab will not be available if the workspace has been configured to display a job-only view, as this view restricts access to the EKM repository.

      • In a multi-region deployment where the EKM administrator has configured EKM to run jobs in a distributed environment, the Repository tab will always be accessing the repository on the Master EKM server, even when you are signed in to a Slave workspace. EKM will then upload the selected files from the Master repository to the Job Data Directory configured on the Slave server.

        For more information, see Running Jobs in a Slave Workspace.


    • To upload files from another location on the EKM server, select the Server tab. Click Add to display a pop-up search window. Navigate to the folder where the files reside, or enter the path to the folder in the address field and click Go. To move up a level in the folder structure, click the up arrow next to the Go button. Select the files that you want to upload by enabling their check boxes, then click OK to add the selected files to the upload list.

  5. Click Next. You can now specify the job submission settings.

    Additional settings may be available if the job was launched from a custom or built-in job template.

    The standard job submission settings are described below.

    Job template

    If you did not launch the job from a job template you can select one now. Templates save time by automatically filling in the job settings for you, and provide a way to specify variables that have been defined in the template. See Starting Jobs in EKM for details.


    Note:
    • The Job template drop box only displays templates saved in Home/My Applications and Administration/Shared Applications unless you started the job directly from a job template that is located in a different folder. In this case the template from which you launched the job is listed under Other Templates in the Job template drop box. However, any other templates saved outside of My Applications and Shared Applications are not displayed.

    • If a job template is selected, settings for command line arguments, batch submission options and the transcript file are not available. These settings are defined in the template.


    Application

    The application that you want to use for computing results, such as ANSYS Workbench, Fluent, or Mechanical APDL.

    Queue

    The job submission queue. You may want to select a specific queue if the application exists on multiple queues.


    Note:  Only MAPDL/Mechanical jobs support the use of a local scratch directory. For all other job types, make sure that you select a queue that uses a shared staging directory.


    Number of cores

    The number of central processing units to reserve on the server.

    Command line arguments

    Command line arguments for the application.

    Batch submission options

    Optional job submission arguments for non-RSM queue types.

    Transcript file

    The name to be used for the transcript file that will be written during job execution. When viewing a job on the Job Monitor page, the transcript is displayed on the Transcript tab for monitoring purposes.

    Import files into repository after job completion

    If enabled, files in the job's working directory will be copied to a repository folder after the job has been executed.

    Once enabled:

    1. Click Browse to select an Output folder for the files.

    2. If you would like to import specific files or file types into the repository, specify a comma-separated list of patterns in the File patterns box, as shown in the example below. Otherwise, if you specify no patterns, all files in the working directory will be imported.


    Note:
    • If you have applied a tag to the job, the Output folder will be automatically set to the folder associated with the tag.

    • This option is not available if you are working in a job-only view.

    • You cannot import files that reside within a sub-folder of the working directory. They must be at the working directory level.


    View command line

    To view the full command line, which includes the execution path, click the View command line link. The command line is displayed in a pop-up window:

  6. Click Finish to execute the job. The status of the job is displayed on the Jobs/Job Monitor page. Provided that the job is not interrupted, the status will go from Queued to Running, and then to Finished once complete.

When executing a batch job you can access the job’s working directory and perform various data management tasks such as uploading, copying, moving, deleting, and renaming files. See Accessing a Job's Working Directory for details.


Note:
  • When you execute a job, an ekm_job.out file is created in the job’s working directory. If you plan to execute a job again, make sure that this file is not being viewed at that time. EKM cannot generate a new ekm_job.out file if the existing one is being viewed, and will report an error if that is the case.

  • A maximum of 10MB of data can be read from the ekm_job.out file. If the size of this file exceeds 10MB, only a partial transcript will be displayed on the Transcript tab.

  • The ekm_job.out file does not include printouts from the solver to RSM’s stderr file.

  • Variable values that you have specified in the Start Job dialog box will be saved in memory when you execute the job. If you execute the job again, the previously entered values will be displayed in the Start Job dialog box for your convenience.



Release 18.2 - © ANSYS, Inc. All rights reserved.