Skip to content

Grafana dashboards and alerts#

This page describes how to install and configure Bufstream's Grafana dashboards and alerts.

Downloads#

Dashboard Description
bufstream-grafana-v0.3.31.zip Bufstream dashboard .json and alert .yaml files for Grafana

About the dashboard#

The overview dashboard (grafana-overview-dashboard.json) provides an Overview row for high-level Bufstream status and rows of panels for Bufstream brokers, Kafka Producers/Consumers, Bufstream metrics (including data governance), and etcd metrics.

Top-level label drop-downs allow filtering by the Kubernetes cluster, Bufstream cluster, Kafka API key, topic, and consumer group.

Bufstream Overview Dashboard

About the status alert#

The status alert (prometheus-status-rules.yaml) fires when any Bufstream broker's status probes report alerts. If any Bufstream broker is in an unhealthy status, it reports an alert to the status probe. This monitor is based on the bufstream.status metric.

Installation#

Dashboard#

To import the dashboard, follow Grafana's instructions, using dashboard-overview.json.

The dashboard exposes its overall health status to aid in identifying and diagnosing operational issues.

Alerts#

Add the rules in prometheus-status-rules.yaml to your deployment, which should look something like this:

serverFiles:
  alerting_rules.yml:
    # rules from prometheus-status-rules.yaml

Configuration#

When using Grafana, Bufstream metrics can either be scraped at a Prometheus endpoint or pushed to a collector.

Prometheus#

If Bufstream is configured to expose a Prometheus endpoint, metrics can be scraped at https://0.0.0.0:9090/metrics/.

We recommend setting pod annotations for the Bufstream pods. Specify the annotations in your Helm values file:

bufstream:
  deployment:
    podAnnotations:
      prometheus.io/scrape: 'true'
      prometheus.io/path: '/metrics'
      prometheus.io/port: '9090'

OTLP#

If you've configured metrics to be exported in OTLP, an OLTP-compatible collector such as Alloy is also needed.

etcd#

To monitor etcd, you may need to enable metrics for your etcd deployment. For etcd dashboards, see etcd's recommended template.

Release notes#

v0.3.6#

Release Date: 2025-01-14

  • Update dashboards to use new fetch and produce request metrics instead of previous errors metrics.

v0.3.3#

Release Date: 2024-12-19

  • Initial release of Grafana dashboard and alerts for Bufstream.