DataMuru Documentation

Current documented release: 0.5.1a0 alpha

Python-first data infrastructure and governance, planned before it changes anything.

Start locally Check capabilities

safe evaluation loop

datamuru validate --strict
datamuru doctor
datamuru plan --target catalog:analytics

DataMuru is a provider-agnostic declarative data infrastructure and governance framework. You describe the platform resources, access rules, and governance intent that you want. DataMuru validates the configuration, compares it with state and supported live resources, shows a deterministic plan, and applies approved changes.

Databricks is the first live provider adapter in the current alpha. The product direction is broader: a shared control layer for data platform resources, governance, brownfield adoption, and eventually multi-cloud execution.

Safe to evaluate today

• initialize projects;
• validate configuration;
• plan changes before mutation;
• apply local state changes;
• connect to Databricks in read-only mode;
• manage selected catalogs, schemas, and grants;
• discover and import supported existing resources for review.

Not a full replacement yet

• general Terraform replacement;
• every Databricks object type;
• production cloud state backends;
• broad multi-workspace orchestration;
• live enforcement for every governance policy;
• transactional rollback guarantees.

pip install "datamuru==0.5.1a0"
datamuru validate --config datamuru.yml
datamuru doctor --config datamuru.yml
datamuru plan --config datamuru.yml

Alpha software

DataMuru 0.5.1a0 is an alpha release. It supports real Databricks operations for the resource types listed in the capability reference, but it is not yet a complete production platform. Test with non-production resources first.

Choose your goal

Goal	Start here
Decide whether DataMuru fits your team	Evaluation checklist
Evaluate DataMuru without cloud access	Five-minute local quickstart
Connect a Databricks workspace safely	Connect a Databricks workspace
Try Databricks Free Edition	Databricks Free Edition
Create a catalog and schemas	Provision a catalog and schemas
Plan an enterprise pilot	Enterprise rollout playbook
Adopt existing Databricks resources	Import an existing workspace
Look up a command or field	CLI reference
Diagnose a failure	Troubleshooting
Evaluate product and edition scope	Platform overview

Who this documentation is for

Reader	What to read first
New evaluator	Choose a path, then Evaluation checklist.
Platform engineer	How DataMuru works, Lifecycle model, and Review and apply changes.
Databricks operator	Authenticate to Databricks, Execution modes, and Catalogs and schemas.
Governance owner	Governance model, RBAC model, and ACL guidelines.
Library contributor	Architecture overview, Library architecture, and Command lifecycle.
Enterprise pilot team	Product requirements summary, Enterprise rollout playbook, and Enterprise testing runbook.

What you can manage today

The OSS alpha includes:

• configuration loading and semantic validation;
• local state, deterministic plans, targeted operations, and saved plans;
• live Databricks catalog and schema reconciliation;
• catalog creation with a managed location or Databricks default storage;
• Unity Catalog grants compiled from DataMuru RBAC definitions;
• read-only discovery and YAML generation for existing workspaces;
• taxonomy, RBAC, and masking definitions;
• a Python API over the same engine used by the CLI.

How the docs are organized

• Start helps you choose a safe evaluation path.
• Tutorials walk through complete tasks from blank project to live provider behavior.
• How-to guides answer specific operator questions.
• Concepts explain the mental model behind state, lifecycle, targets, and governance.
• Architecture explains the package design, command lifecycle, provider contract, and extension points.
• Reference defines command, config, result, and error contracts.
• Operations covers team adoption, release, security, production readiness, and enterprise rollout practices.
• Product explains edition boundaries, requirements, roadmap, packaging, and enterprise testing.

Some resource types are modeled locally but do not yet have live provider effects. Account-level identity lifecycle is an Enterprise capability and also depends on Databricks account SCIM availability. See Current capabilities and limits before planning a rollout.

The operating loop

1. Write configuration. Keep project, provider, workspace, environment, and governance concerns in separate YAML files.
2. Validate. Catch missing files, unsupported values, and unsafe edition combinations before contacting a provider.
3. Run doctor. Verify credentials, connectivity, SQL warehouse requirements, and provider capabilities.
4. Plan. Review create, update, no-op, and destroy actions.
5. Apply deliberately. Prefer narrow targets during evaluation and saved plans in shared environments.
6. Re-plan. A successful idempotent run should report no required changes.

Documentation conventions

• Commands use datamuru, the console script installed from PyPI.
• Paths are relative to the directory containing datamuru.yml unless stated otherwise.
• Placeholder values use names such as your-workspace and must be replaced.
• Destructive and live-cloud steps are marked explicitly.
• Every task states prerequisites, expected results, and a recovery path.

Get help

Search this site first. If the problem remains:

• review Troubleshooting;
• collect datamuru doctor --output json and the structured error code;
• choose the appropriate route in Support and feedback without including tokens, customer data, or private workspace details.