Scheduled input agents (cron)

Scheduled input agents fetch data from a data source at a user-defined time. You specify the time in classic cron syntax or in a simplified form. You can also trigger the input agent via HTTP or message.

Notes

• The Control Center shows the last and the next scheduled start time of a profile.

• You can create directory names dynamically with a date in all data sources. For example, on August 23, 2023, the system replaces the directory name C:/tmp/<yyyyMMdd> with C:/tmp/20230823.

• The system variable VAR_SYS_TRIGGER_CHANNEL_ID (or MSG_CALL_VAR_SYS_TRIGGER_CHANNEL_ID) replaces the channel in the scheduled input agent. This works with every data source. As the value of the variable, specify the ID of the new channel.

• A job might not end normally after a forced stop or a power failure. In that case, the system marks the job as a "crashed job." It restarts the job at the next server start. The restart begins from the start, not from the point of interruption. This requires trackCrashedJobs = true in startup.xml.

  Node Controller / Load Balancing: By default, cron jobs always run on the Node Controller, not on clients. You can assign a specific node per profile.

Creating internal cron jobs with a function

See the function createScheduleEntry().

Cron job for deleting old files

See section Deleting files with internal cron job.