There are several areas within LityxIQ where a process (such as a Dataset or Model execution) can be put on a schedule. In most cases, you will see the "Schedule It" option or something similar as one of the options available for executing that object. For example, in the case of a Raw Dataset, the Load Data option expands into options Load Now or Schedule It.
Regardless of the object, a common scheduling dialog will appear. Here we'll explain the components of that dialog and the options available.
Settings - Left Side of Scheduling Dialog
Schedule Type:
- Disable Scheduling - select this option to disable scheduling for the object. By default, all objects have scheduling disabled when created.
- By Date - select this option to use a date-based schedule, such as daily or weekly. See below for more detail.
- Upon Data Refresh - select this option to have an object updated automatically, when its source information is updated. The exact meaning of this is different for different objects. For example, in the case of a predictive model, this setting will automatically re-build the model whenever the dataset feeding the model is updated. See the relevant knowledge base document for the meaning specific to the object type you wish to schedule.
Repeat Settings: If you select a "By Date" schedule type, you must also set when and how often it repeats.
- Does Not Repeat - select this option to schedule just a single future run.
- Custom Number of Days - run the object on a specific periodic interval, every given number of days. The number of days is entered in the Repeat Every box on the right side of the dialog.
- Custom Number of Hours - run the object on a specific periodic interval, every given number of hours. The number of hours is entered in the Repeat Every box on the right side of the dialog.
- All other options (e.g., Hourly, Weekly, Monthly) will set the object to run once per the time period selected. For example, if you select "Weekly", the object will run on the selected start date/time, and then each one week after that until the end date is reached.
Error Action Selection: This drop down is used to specify the action the program should take in the event that an error occurs on a scheduled run.
- Cancel Future Runs - with this setting, if an error occurs in a schedule run, the object will not run again automatically. You can still run it again manually, and if eventually successful, the scheduled runs will continue again afterward.
- Continue Runs on Schedule - with this setting, the object will continue attempting to run on schedule even in the case of errors. In this case, after an error occurs, the process will attempt to execute again every ten minutes. Note that future runs are often likely to continue producing errors and so it is a good idea to investigate any underlying issues when an error occurs. However, this setting is a good choice if temporary errors might occur (such as database connectivity to an external data source being temporarily interrupted) that are likely to resolve themselves naturally so that future LityxIQ runs will be successful.
Dates/Times - Right Side of Scheduling Dialog
These options are available only if a Date-based schedule was selected.
Start Date/Time: Select the date and time for the initial run. This setting is interpreted in your current timezone setting.
Repeat Every: If the "Custom Number of Days" or "Custom Number of Hours" setting were chosen, this entry become available to specify the number of days or hours between executing the object.
End Date: Select the final date for scheduled runs. Runs will not be executed after this date. Note that this option is not displayed for Does Not Repeat schedules.
To save the schedule and exit the dialog, click Save, or to cancel your changes, click Cancel.