Schedulers Endpoint

Overview

The Scheduler actively takes data from Elasticsearch and sends it for analysis to the Engine API.

The Scheduler is primarily designed for running analysis in near real-time. As soon as data becomes available in Elasticsearch is can be analyzed. It is also possible to use the Scheduler to analyze historical data over a fixed time range.

Data may be pulled from the same Elasticsearch instance as the Engine API or from a separate instance. The Elasticsearch HTTP REST API is used which is supported by Elasticsearch 1.7 and above.

Configuring the Scheduler

The configuration must be supplied when the job is created. This is easiest to do using the Prelert Kibana plugin.

When starting the Scheduler, you may specify a start time. This allows you to include a training period, providing you have this data available in Elasticsearch. If you wish to analyze from the beginning of a dataset, you may specify any date earlier (minimum Jan 01 1970).

When starting the Scheduler, you may specify an end time. If specified, the Scheduler will analyze data from the start time until the end time at which point the scheduler will stop. This is useful for a one-off batch analysis.

If you do not specify an end time, the Scheduler will run continuously.

Upon a system restart, any jobs where the Scheduler was running will also be restarted.