Skip to content

Latest commit

 

History

History
executable file
·
71 lines (56 loc) · 3.1 KB

README.md

File metadata and controls

executable file
·
71 lines (56 loc) · 3.1 KB
Raw

Customize your cards

While prometheus-msteams comes with a default card template out of the box, you can easily provide a custom template file to adapt your notification cards to your exact requirements. If you have a template file you would like to contribute to this collection of examples don't hesitate to open a pull request or issue.

Table of Contents

Silence Alerts from MS Teams

A silence action card template has been created for those who need acknowledge the alerts from MS Teams. The template works together with Alertmanger

silencecard

This template provides an action which opens the Alertmanager UI with all the information about your alert and the Silence is created via Alertmanager.

silencepreview

Grafana Alert like Card Template

This template shows keys and values in a list similar to how Grafana formats alerts for Teams.

screenshot_card_template_grafana_group

When you use the MS Teams integration for your Grafana Alert, the notifications are grouped by alert while also listing the legend and value of every time series that matched the respective alert expressions. This template tries to recreate and improve this behavior and card layout.

It comes with a number of requirements alert rules must fullfill. Therefore the template also contains a fallback template that is more or less the same as the default provided with prometheus-msteams. To trigger the grafana layout the annotation template must be set to "group". Your Alertmanager must also be configured to group at least by alert name.

A detailed info on all requirements including examples can be found as a comment in the template file itself.

- alert: Memory_Ratio_WorkingSetAndRss_Reservation_High
  expr: |
    max by (namespace, name) (
      (
        avg_over_time(container_memory_working_set_bytes[1h])
        > avg_over_time(container_memory_rcontainer_memory_rss[1h])
        or avg_over_time(container_memory_rcontainer_memory_rss[1h])
      )
      / on (namespace, name, id) avg_over_time(container_spec_memory_reservation_limit_bytes[1h]) != +Inf
    ) * 100 > 150
  for: 12h
  labels:
    severity: "info"
    alert_with_resolve: "true"
  annotations:
    title: "Memory: Relativ memory usage above reservation"
    summary: "Ratio(Max(Working Set, RSS) / Reservation) [1h Avg for 12h] > 150 %"
    description: >
      The ratio of the working set memory or resident set size against the
      configured reservation has been above 150 % percent as a 1h average
      for more than 12h.
    key: "{{ $labels.name }} | {{ $labels.namespace }}"
    value: '{{ $value | printf "%.2f" }} %'
    investigate_link: 'https://${DOMAIN}/promstack/grafana/d/fdwTIxFMz/container-intra'
    template: "group"