The Scheduler
You can use the Scheduler to create jobs to run administrative tasks or collect information at predictable times. The Scheduler uses SQL statements instead of operating system job scheduling tools.
The Scheduler is controlled by a set of tables in the sysadmin database.
- Task
- Runs an action at a specific time and frequency.
- Sensor
- Runs an action at a specific time and frequency to collect data, create a results table, store the data in the results table, and purge old data after a specified time.
- Startup task
- A task that runs only when the server moves from quiescent mode to online mode.
- Startup sensor
- A sensor that runs only when the server moves from quiescent mode to online mode.
The action of a task or sensor can be one or more SQL statements, user-defined routines, or stored procedures.
- Associate tasks and sensors into functional groups
- Track the execution time and return value each time a task or sensor is run
- Define alerts with varying severity
- Define thresholds to control when tasks or sensors are run
The Scheduler contains built-in tasks and sensors that run automatically. You can modify the built-in tasks and sensors and define your own tasks and sensors.
Disk space requirements
The Scheduler tables and sensor results tables can use significant amounts of disk space.You can use the following formula to estimate the disk usage for one sensor:
Number of rows collected * size of the row collected * the frequency of data collection per day * the retention period
Repeat this estimate for all sensors and you can determine a close estimate of the space required.
You can reduce the amount of data stored by decreasing the frequency of data collection or shortening the retention period by updating the ph_task table.
You can move the sysadmin database to a different dbspace by using the SQL administration API, however, all existing data in the database will be lost.