OpenNMS Helm - PM/FM Console for Grafana

Helm is a Grafana application that allows users to create flexible dashboards using both fault management (FM) and performance management (PM) data from OpenNMS® Horizon™ and/or OpenNMS® Meridian™.

It is the successor to the original OpenNMS Datasource for Grafana.

Consult the Helm Documentation for help on installing, configuring and using the application.

Features

Fault Management (FM)

Flexible alarm filtering

Isolate the alarms you want displayed using custom filters and conditions that can match against over 150 different fields.

Configurable displays

Present the fields that are most relevant to you and your teams for improved tracking and triage.

Alarm interactions

Acknowledge, escalate and create tickets for alarms directly from the dashboard.

Performance Management (PM)

Storage engine agnostic

Retrieve metrics stored in any of the available persistence engines i.e. rrdtool, JRobin and Newts.

Template support

Populate template variables with all of the nodes belonging to a set of categories, or leverage the complete filter grammar.

Trending and Forecasting

Remove outliers and perform trending or forecasting using the built-in series filters. Develop your own filters using Java or R.

Derive new series using JEXL expressions.

Support Matrix

• Performance Management Data Source
  • OpenNMS Horizon v16.x or greater
  • OpenNMS Meridian v2016.1.0 or greater
• Fault Management Data Source
  • OpenNMS Horizon v20.1.x or greater
  • OpenNMS Meridian v2017.1.0 or greater

Issue Tracking

We use the Helm project in our JIRA instance to track bugs and enhancements related this to project.

Changelog

v3.0.0

General

• Improved error messages for incomplete or invalid queries

Performance Data Source

• Added support for overriding time intervals and max datapoints
• Labels are now shown in the order they were queried
• Labels can now be formatted using transformation functions like nodeToLabel(<foreignSource:foreignId>) and resourceToName(<resourceId>) (Horizon 24 or greater)
• Measurements API requests are now made in relaxed mode (if the server is missing a particular requested attribute, all others are still returned)

Flow Data Source

• Added additional transforms for flow data (toBits, onlyIngress, onlyEgress)
• The Flow Deep Dive dashboard axis labels have been updated to be more intuitive

Fault Management Data Source

• Sorting by numeric columns now works as expected
• HTML alarm (event) log messages are now rendered properly
• Alarm multi-select and deselect now works as expected
• "Severity" in the Alarm Table panel is now a normal column, rather than a "Severity icons" check box in the config options -- existing configs should be automatically upgraded
• Added support for reordering columns in the Alarm Table panel
• Added support for Situations (correlated alarms), including sending feedback on alarm correlations (Horizon 23 or greater)
• Added limited support for multi-select dashboard variables in the Alarm Table panel
• Added a custom node attribute in the Alarm Table panel that supports node criteria (either passing a nodeId or foreignSource:foreignId tuple)
• Many enhancements were made to the Alarm Detail view (full raw response viewing, related alarms, etc.)
• It is now possible to display dates as a relative time in the Alarm Table panel

v2.0.0

• Added a new datasource for querying flow data from OpenNMS
• Added support for "fallback" attributes to the performance datasource
• Added the ability to configure query timeouts for all of the datasources
• Require Grafana 5.x or greater

v1.1.0

• Added support for custom actions in the Alarm Table panel
• Added the operator instructions field to the alarm details modal
• Updated the package dependencies to support Grafana 5.x
• Fixed a bug where long alarm descriptions and log messages would fill the alarm details modal
• Use consistent colors in both the Alarm Table and Alarm Histogram panels

v1.0.0

• Initial release