Adding and configuring job servers

An analysis launched by clicking Start Analysis in the Study Tasks pane runs on the current machine. You can also use the current machine to launch an analysis on another machine on the network, provided the same release of Autodesk Simulation Moldflow Insight - Solvers is installed on the other machine, and you have created a Job Server entry for the other machine in the Job Manager dialog.

Tip: The Job Server feature of the Job Manager is the most convenient way to run analyses on a solver-only platform.

A default job server is setup during installation, and assigned the name localhost. If you change this default name, localhost will not appear in your job manager list when you choose a job server on the network.

To add a new job server:

1. Click (Home tab > Analysis panel > Job Manager).
2. Click Add Job Server to open the Job Server Properties dialog.
3. In the Server address text box, enter the address of the machine on which you want the analyses to run. You can specify the address using one of the following three methods:

   IP address

   For example, 190.11.222.333

   Host name

   For example, john-smith

   Fully qualified host name

   For example, john-smith.mycompany.com

4. In the Displayed name text field, enter the job server name you want to appear in the Job Manager dialog.
5. In the Batch queue max. running jobs text field, enter the maximum number of simultaneous batch jobs that are allowed to run on the current job server. You should consider the size of the models you will be analyzing, the number of CPUs in the job server machine, and the available memory and disk space resources for the job server machine when selecting this limit.
   Note: Jobs submitted as Priority Jobs take precedence over jobs submitted to the batch queue and reduce the number of jobs that can run simultaneously in the batch queue.

   If a job reaches the top of the queue, and requires a licence which is currently reserved, the job manager will wait until the license becomes unreserved.

6. If you want the jobs submitted to the batch queue to start at a particular time, select the Start queue at option and specify the date and time.
7. Select the desired Default job priority level setting. The priority at which a job runs affects how intensively the job server's resources are used. The default setting of Low ensures that other programs running on the machine will be allocated adequate CPU resources while your analyses are running. If no other programs are running on the machine, the priority setting will not affect the speed of the analysis.
   Note: It is not recommend to set the priority to Highest on a single CPU machine, as this can make the system unresponsive. The priorities Low and Lowest are equivalent on a Windows system.
   Tip: If two analyses are running on the same CPU, you can "speed up" one of the analyses relative to the other by increasing that job's priority via the context menu of the analysis entry in the Job Manager.

If you will be running Design-Of-Experiments, Runner Balance or Process Optimization analyses, which launch a series of studies (experiments) and collate the results, you can optionally set up a pool of job servers for those analyses to run on. In the Job distribution pool list, select the job servers that you want the experiments to run on. You can monitor the status of this job distribution in the Distribution Queue.

Note: To display the Distribution Queue with IPv6 networks, you must use the string localhost to specify the local machine.

If desired, you can change the directory in which the temporary files of an analysis running on the job server are created, and change the amount of time that temporary files are retained in that directory before they are automatically deleted. The default setting is 7d, which means seven days.