Prometheus Filter (4.0)

With the Prometheus filter you can configure the Prometheus metrics that are going to be exposed for scraping.

The purpose of the Prometheus Filter is to prevent the flooding of metrics in the storage of the Prometheus host.

The filter is positive, so you define what you want to get. Except in the situation where no Prometheus Filter exists or all existing Prometheus Filters are disabled, then all available metrics are exposed for scraping. The exception is MIM metrics from batch workflows. In the case of MIMs from batch workflows, no metrics will be collected unless there is a filter whitelisting them.

A Prometheus filter consists of one or many metric filters. Each metric filter has a Metric Name Filter and optionally a number of Label Filter(s). A metric that passes at least one of the metric filters will be exposed for scraping.

Open

Example of a Prometheus filter with one metric filter that has no label filters

To Add a Metric Filter

To add a metric filter, take the following steps:

1. Click the Add button. 

2. In the dialog that is displayed, enter a Metric Name Filter in the form of a regular expression. 

3. Optionally, specify one or many label filter(s) by entering the Label along with a corresponding Label Value Filter
   The Label Value Filter has to be entered in the form of a regular expression. 

4. Click Add to add the metric filter.

5. Click Close when done. 

Open

For example, five metric filters are configured. See the detailed description in the table below.