Skip to main content

Database Drift Detection & Schema Monitoring

Atlas Schema Monitoring tracks the schema of your live databases and reports it to the Atlas Cloud control plane, either from an agent (container) in your database VPC or from a scheduled CI/CD pipeline. Using this metadata, Atlas provides:

  1. Live visibility of your database schema with automated ER diagrams and auto-generated documentation.
  2. A Changelog of all schema changes, so you can see how schemas change over time, and easily triage schema change related issues.
  3. Schema statistics: Get insights into your database schema with statistics like table and index sizes.
  4. Alerts: Use Slack notifications to inform or alert teams that need to know about schema changes or drifts.

GitHub Actions Quickstart

Get visibility into your database schema in under 5 minutes using GitHub Actions.

GitLab CI/CD Quickstart

Get visibility into your database schema in under 5 minutes using GitLab CI/CD.

BitBucket Pipelines Quickstart

Get visibility into your database schema in under 5 minutes using Bitbucket Pipelines.

Agent Quickstart

Get visibility into your database schema in under 5 minutes using the Atlas agent.

Helm

Deploy the Atlas agent using Helm on Kubernetes.

Terraform (AWS)

Deploy the Atlas Agent to AWS using Terraform.

Features​

Drift Detection

Automatically monitor schema drifts.

Database Discovery

Automatically discover database instances and start monitoring them with the Atlas agent.

Monitoring as Code

Configure Atlas Agent monitoring using HCL configuration files for version-controlled, declarative monitoring.

Slack Notifications

Send schema change and drift alerts to Slack channels.

How It Works​

Architecture

Learn about the key concepts and features of Atlas Database Schema Monitoring.

Database Credentials and Security

Discover how to securely provide database credentials to the Atlas agent.

What features are essential in a schema monitoring tool?​

A schema monitoring tool has to answer five questions about every database it watches:

  1. What is the schema right now? The tool inspects the live database and keeps a current picture of it, including objects such as views, triggers, and functions.
  2. What changed, and when? Changes are recorded in a changelog, so an incident can be traced back to the schema change that caused it.
  3. Does the database still match the source of truth? Drift detection compares the live schema against the migration directory or schema repository and reports the difference.
  4. Who gets told? Slack notifications push change and drift events to the team that owns the database, instead of waiting for someone to open a dashboard.
  5. How big is it, and how is it growing? Alongside schema snapshots, the agent collects table statistics for every monitored database, configurable per database or schema.

Two implementation details matter as much as the feature list: the tool should only read schema metadata and never the data in your tables, and it should run inside your network instead of asking you to open the database to a vendor. Atlas does both. The agent runs in your environment and connects to the database directly, with credentials that never reach the control plane, or a GitHub Action reports the same metadata from your CI pipeline.