Atlas: manage your database schema as code

Atlas is a language-agnostic tool for managing and migrating database schemas using modern DevOps principles. It offers two workflows:

• Declarative: Similar to Terraform, Atlas compares the current state of the database to the desired state, as defined in an HCL, SQL, or ORM schema. Based on this comparison, it generates and executes a migration plan to transition the database to its desired state.

• Versioned: Unlike other tools, Atlas automatically plans schema migrations for you. Users can describe their desired database schema in HCL, SQL, or their chosen ORM, and by utilizing Atlas, they can plan, lint, and apply the necessary migrations to the database.

Quick installation

macOS + Linux:

curl -sSf https://atlasgo.sh | sh

Homebrew:

brew install ariga/tap/atlas

Docker:

docker pull arigaio/atlas

NPM:

npx @ariga/atlas

Click here to read instructions for other platforms.

Getting started

Get started with Atlas by following the Getting Started docs. This tutorial teaches you how to inspect a database, generate a migration plan and apply the migration to your database.

Key features:

• Schema management: The atlas schema command offers various options for inspecting, diffing, comparing, and modifying database schemas.
• Versioned migration: The atlas migrate command provides a state-of-the-art experience for planning, linting, and applying migrations.
• Terraform support: Managing database changes as part of a Terraform deployment workflow.
• SQL, HCL and ORM support: Atlas enables users to define their desired database schema using HCL, SQL, or their chosen ORM.
• Multi-tenancy: Atlas includes built-in support for multi-tenant database schemas.
• Cloud integration: Atlas integrates with standard cloud services and provides an easy way to read secrets from cloud providers such as AWS Secrets Manager and GCP Secret Manager.

schema inspect

Easily inspect your database schema by providing a database URL and convert it to HCL, JSON, SQL, ERD, or other formats.

Inspect a specific MySQL schema and get its representation in Atlas DDL syntax:

atlas schema inspect -u "mysql://root:pass@localhost:3306/example" > schema.hcl

Result

table "users" {
  schema = schema.example
  column "id" {
    null = false
    type = int
  }
  ...
}

Inspect the entire MySQL database and get its JSON representation:

atlas schema inspect \
  --url "mysql://root:pass@localhost:3306/" \
  --format '{{ json . }}' | jq

Result

{
  "schemas": [
    {
      "name": "example",
      "tables": [
        {
          "name": "users",
          "columns": [
            ...
          ]
        }
      ]
    }
  ]
}

Inspect a specific PostgreSQL schema and get its representation in SQL DDL syntax:

atlas schema inspect \
  --url "postgres://root:pass@:5432/test?search_path=public&sslmode=disable" \
  --format '{{ sql . }}'

Result

-- create "users" table
CREATE TABLE "users" ("id" integer NULL, ...);
-- create "posts" table
CREATE TABLE "posts" ("id" integer NULL, ...);

Inspect a specific PostgreSQL schema and get its ERD representation in the browser:

atlas schema inspect \
  --url "postgres://root:pass@:5432/test?search_path=public&sslmode=disable" \
  -w

Inspect a specific PostgreSQL schema and get its ERD representation Mermaid syntax:

atlas schema inspect \
  --url "postgres://root:pass@:5432/test?search_path=public&sslmode=disable" \
  --format '{{ mermaid . }}'

erDiagram
    users {
      int id PK
      varchar name 
    }
    blog_posts {
      int id PK
      varchar title 
      text body 
      int author_id FK
    }
    blog_posts }o--o| users : author_fk

Loading

schema diff

Compare two schema states and get a migration plan to transform one into the other. A state can be specified using a database URL, HCL or SQL schema, or a migration directory.

Compare two MySQL schemas:

atlas schema diff \
  --from mysql://root:pass@:3306/db1 \
  --to mysql://root:pass@:3306/db2

Result

-- Drop "users" table
DROP TABLE `users`;

Compare a MySQL schema with a migration directory:

atlas schema diff \
  --from mysql://root:pass@:3306/db1 \
  --to file://migrations \
  --dev-url docker://mysql/8/db1

Compare a PostgreSQL schema with an Atlas schema in HCL format:

atlas schema diff \
  --from "postgres://postgres:pass@:5432/test?search_path=public&sslmode=disable" \
  --to file://schema.hcl \
  --dev-url "docker://postgres/15/test"

Compare an HCL schema with an SQL schema:

atlas schema diff \
  --from file://schema.sql \
  --to file://schema.hcl \ 
  --dev-url docker://postgres/15/test  

schema apply

Generate a migration plan and apply it to the database to bring it to the desired state. The desired state can be specified using a database URL, HCL or SQL schema, or a migration directory.

Update the database to the state defined in the HCL schema:

atlas schema apply \
  --url mysql://root:pass@:3306/db1 \
  --to file://schema.hcl \
  --dev-url docker://mysql/8/db1

Result

-- Planned Changes:
-- Modify "users" table
ALTER TABLE `db1`.`users` DROP COLUMN `d`, ADD COLUMN `c` int NOT NULL;
Use the arrow keys to navigate: ↓ ↑ → ← 
? Are you sure?: 
  ▸ Apply
    Abort

Update the database to the state defined in a specific version of the migration directory:

atlas schema apply \
  --url mysql://root:pass@:3306/db1 \
  --to "file://migrations?version=20221118091226" \
  --dev-url docker://mysql/8/db1

Additional schema commands

Atlas offers additional commands to assist users managing their database schemas. These include schema clean and schema fmt. For more information, see the versioned migration documentation at https://atlasgo.io/declarative/inspect.

migrate diff

Write a new migration file to the migration directory that bring it to the desired state. The desired state can be specified using a database URL, HCL or SQL schema, or a migration directory.

Create a migration file named add_blog_posts in the migration directory to bring the database to the state defined in an HCL schema:

atlas migrate diff add_blog_posts \           
  --dir file://migrations \
  --to file://schema.hcl \
  --dev-url docker://mysql/8/test

Create a migration file named add_blog_posts in the migration directory to bring the database to the state defined in an SQL schema:

atlas migrate diff add_blog_posts \           
  --dir file://migrations \
  --to file://schema.sql \
  --dev-url docker://mysql/8/test

Create a migration file named add_blog_posts in the migration directory to bring the database to the state defined by another database:

atlas migrate diff add_blog_posts \           
  --dir file://migrations \
  --to mysql://root:pass@host:3306/db \
  --dev-url docker://mysql/8/test

migrate apply

Apply all or part of pending migration files in the migration directory on the database.

Apply all pending migration files in the migration directory on a MySQL database:

atlas migrate apply \
  --url mysql://root:pass@:3306/db1 \
  --dir file://migrations

Apply in dry run mode the first the pending migration file in the migration directory on a PostgreSQL schema:

atlas migrate apply \
  --url "postgres://root:pass@:5432/test?search_path=public&sslmode=disable" \
  --dir file://migrations \
  --dry-run

Additional migrate commands

Atlas offers additional commands to assist users managing their database migrations. These include migrate lint, migrate status, and more. For more information, see the versioned migration documentation at https://atlasgo.io/versioned/diff.

Supported databases

MySQL, MariaDB, PostgresSQL, SQLite, TiDB, CockroachDB, SQL Server, ClickHouse, Redshift.

Supported Version Policy

To ensure the best performance, security and compatibility with the Atlas Cloud service, the Atlas team will only support the two most recent minor versions of the CLI. For example, if the latest version is v0.25, the supported versions will be v0.24 and v0.25 (in addition to any patch releases and the "canary" release which is built twice a day).

Old Binaries

As part of the Supported Version Policy mentioned above, binaries for versions that were published more than 6 months ago will be removed from the CDN and Docker Hub.