ServicesAboutNotesContact Get in touch →
EN FR
Note

Elementary alerting hub

A reading path through Elementary's alerting system -- from the edr monitor command through Slack/Teams setup, filter-based routing, alert fatigue reduction, and on-call strategy.

Planted
dbtelementarydata qualityautomation

Elementary stores test results in the warehouse. The alerting layer — edr monitor and its configuration — routes those results to the right people. These five notes cover the full alerting stack, from the basic command through the operational decisions about who gets paged and when.

Reading order

  1. Elementary edr monitor alerting — The edr monitor command: how it works, how it differs from edr report, and how to configure alert metadata (owner, channel, suppression interval) in model YAML. Start here before configuring any destination.

  2. Elementary Slack and Teams integration — Token-based vs webhook Slack integration (and why token-based is worth the extra setup), per-model and path-based channel routing, and the current state of Microsoft Teams with the incoming webhook deprecation.

  3. Elementary alert routing with filters — Running multiple edr monitor commands with different --filters to route by tag, owner, status, or resource type. Includes the PagerDuty bridge pattern for OSS users who need incident management without upgrading to Elementary Cloud.

  4. Elementary alert fatigue reduction — Suppression intervals, alert grouping, threshold configuration, sampling controls for PII tables, and what Elementary Cloud adds with automatic incident grouping and resolution.

  5. Data team on-call strategies — The human side: triage severity frameworks, embedding runbooks in test metadata, MTTA/MTTR as diagnostic metrics, and rotation patterns that account for how data teams actually work.

Where this fits in the broader stack

These notes assume Elementary is installed and has populated test results. If you’re starting from scratch, the Elementary for dbt note covers installation, anomaly detection configuration, and the dbt package architecture.

The Data Observability Minimum Viable Stack note explains why alerting is one of the four non-negotiable observability capabilities — tests that fail silently provide no protection. And if your team has grown to the point where OSS Elementary alerting is hitting its limits, the Data Observability Build vs. Buy hub covers the decision of when paid tools earn their cost.