Skip to main content

Alerts Overview

Alerts are your first line of defense against database failure. In pgpulse, we believe alerts should be actionable and symptom-based.

What Alerts are For​

An alert should notify you when your database is unable to serve users or when it's approaching a critical resource limit. We avoid "noisy" alerting on transient blips that resolve themselves.

Core Principles​

  • Alert on Symptoms: CPU spikes and connection saturation are symptoms of a problem.
  • Minimize Noise: Don't alert on 50% CPU. Save alerts for when the system is actually under strain.
  • Provide Context: Every pgpulse alert includes a link to the relevant dashboard and query insights so you can start debugging immediately.

Delivery Channels​

You can configure alerts to be delivered via:

  • Email: For consolidated daily summaries and non-critical warnings.
  • Slack: For real-time critical alerts that need immediate attention.
  • Webhook: To integrate with your own custom incident management tools.