Integrate deployment metadata service for locking and state by shreyas-goenka · Pull Request #4856 · databricks/cli

shreyas-goenka · 2026-03-26T16:51:50Z

Summary

Integrates the Deployment Metadata Service (DMS) as an alternative backend for deployment locking and resource state management. Gated behind DATABRICKS_BUNDLE_MANAGED_STATE=true.

When enabled:

Locking: Uses server-side versioned locks (with heartbeats) instead of workspace filesystem lock files
State: Reads/writes resource state via the DMS API (ListResources / CreateOperation) instead of local state files
Operations: Reports each resource operation (create, update, delete) inline to the server with resource state

Key implementation details

DeploymentLock interface (lock.go) with two implementations: workspaceFilesystemLock (existing behavior) and metadataServiceLock (DMS)
resolveDeploymentID reads deployment ID from workspace resources.json, or generates a new UUID for fresh deployments (written only after CreateDeployment succeeds)
LoadStateFromDMS populates the state DB from ListResources instead of reading local files
PushResourcesState is a no-op with DMS (state is persisted per-operation to the server)
--plan flag and bind/unbind are not supported with DMS
Heartbeat goroutine keeps the lock alive during long deployments

Test plan

Acceptance tests under acceptance/bundle/dms/ covering: deploy with resource creation, sequential deploys with create/delete, plan + summary, deploy errors, and lock release errors
Unit test for planActionToOperationAction mapping
E2E testing against staging workspace (32/32 passing)

eng-dev-ecosystem-bot · 2026-03-26T17:01:53Z

Commit: 342fef8

Run: 23764962616

	Env	✅pass	🙈skip	Time
✅	aws linux	73	14	0:46
✅	aws windows	73	14	0:32
✅	aws-ucws linux	76	11	0:41
✅	aws-ucws windows	76	11	0:30
✅	azure linux	72	14	0:49
✅	azure windows	72	14	0:39
✅	azure-ucws linux	75	11	0:43
✅	azure-ucws windows	75	11	0:36
✅	gcp linux	73	14	0:42
✅	gcp windows	73	14	0:31

shreyas-goenka · 2026-04-08T12:28:03Z


+		// Report skip actions to the metadata service. On initial registration,
+		// these are recorded as INITIAL_REGISTER operations.
+		if action == deployplan.Skip && b.OperationReporter != nil {


move the initial registration up

shreyas-goenka · 2026-04-10T12:05:23Z

@@ -0,0 +1,6 @@
+Local = true
+Cloud = false


The service needs to roll out to prod before we enable this on cloud.

shreyas-goenka · 2026-04-14T00:37:12Z

+	assert.True(t, ok)
+	assert.Equal(t, tmpdms.VersionTypeDestroy, vt)
+
+	_, ok = goalToVersionType(GoalBind)


support can be added as a followup.

github-actions · 2026-04-15T00:09:27Z

Approval status: pending

`/acceptance/bundle/` - needs approval

Files: acceptance/bundle/dms/add-resources/databricks.yml, acceptance/bundle/dms/add-resources/out.test.toml, acceptance/bundle/dms/add-resources/output.txt, acceptance/bundle/dms/add-resources/script, acceptance/bundle/dms/add-resources/test.toml, acceptance/bundle/dms/deploy-error/databricks.yml, acceptance/bundle/dms/deploy-error/out.test.toml, acceptance/bundle/dms/deploy-error/output.txt, acceptance/bundle/dms/deploy-error/script, acceptance/bundle/dms/deploy-error/test.toml, acceptance/bundle/dms/plan-and-summary/databricks.yml, acceptance/bundle/dms/plan-and-summary/out.test.toml, acceptance/bundle/dms/plan-and-summary/output.txt, acceptance/bundle/dms/plan-and-summary/script, acceptance/bundle/dms/release-lock-error/databricks.yml, acceptance/bundle/dms/release-lock-error/out.test.toml, acceptance/bundle/dms/release-lock-error/output.txt, acceptance/bundle/dms/release-lock-error/script, acceptance/bundle/dms/release-lock-error/test.toml, acceptance/bundle/dms/sequential-deploys/databricks.yml, acceptance/bundle/dms/sequential-deploys/out.test.toml, acceptance/bundle/dms/sequential-deploys/output.txt, acceptance/bundle/dms/sequential-deploys/script, acceptance/bundle/dms/sequential-deploys/test.toml, acceptance/bundle/dms/test.toml
Suggested: @denik
Also eligible: @pietern, @andrewnester, @anton-107, @janniklasrose, @lennartkats-db

`/bundle/` - needs approval

Files: bundle/bundle.go, bundle/deploy/lock/acquire.go, bundle/deploy/lock/deployment_metadata_service.go, bundle/deploy/lock/deployment_metadata_service_test.go, bundle/deploy/lock/lock.go, bundle/deploy/lock/release.go, bundle/deploy/lock/workspace_filesystem.go, bundle/deployplan/plan.go, bundle/direct/bundle_apply.go, bundle/direct/bundle_plan.go, bundle/direct/pkg.go, bundle/env/deployment_metadata.go, bundle/phases/bind.go, bundle/phases/deploy.go, bundle/phases/destroy.go, bundle/statemgmt/resources_json.go, bundle/statemgmt/state_dms.go, bundle/statemgmt/state_pull.go, bundle/statemgmt/state_push.go
Suggested: @denik
Also eligible: @pietern, @andrewnester, @anton-107, @janniklasrose, @lennartkats-db

`/cmd/bundle/` - needs approval

Files: cmd/bundle/utils/process.go
Suggested: @denik
Also eligible: @pietern, @andrewnester, @anton-107, @janniklasrose, @lennartkats-db

General files (require maintainer)

Files: acceptance/bin/print_requests.py, libs/testserver/deployment_metadata.go, libs/testserver/fake_workspace.go, libs/testserver/handlers.go, libs/testserver/server.go, libs/tmpdms/api.go, libs/tmpdms/types.go
Based on git history:

@denik -- recent work in bundle/direct/, libs/testserver/, bundle/statemgmt/

_{Any maintainer (@andrewnester, @anton-107, @denik, @pietern, @simonfaltum, @renaudhartert-db) can approve all areas.

See OWNERS for ownership rules.}

Add server-side deployment locking and state management via the Deployment Metadata Service (DMS), gated behind DATABRICKS_BUNDLE_MANAGED_STATE=true. Key changes: - DeploymentLock interface with factory (DMS or filesystem based on env) - DMS lock: version-based locking with heartbeat, operation reporting - State read/write via ListResources/CreateOperation with per-resource state - withDeploymentLock helper extracts lock boilerplate from deploy/destroy - Temporary DMS client (libs/tmpdms) mirroring future SDK-generated code - Mock DMS server for acceptance tests - 6 acceptance tests covering deploy, destroy, plan, summary, sequential deploys, and adding resources with remote state Co-authored-by: Isaac

…troy phases Co-authored-by: Isaac

LoadStateFromDMS is a state-loading function, not a lock function. Moving it to statemgmt where it belongs alongside other state management code. Co-authored-by: Isaac

When we just created the deployment, LastVersionID is necessarily empty so we can start at version "1" directly. Co-authored-by: Isaac

Co-authored-by: Isaac

Print requests inline in output.txt and clear remaining requests at the end of each script so out.requests.txt is not generated. Also update sequential-deploys test to add/remove resources across deploys, asserting create and delete operations are captured. Co-authored-by: Isaac

…erations - Print DMS requests inline in output.txt via print_requests.py - Update sequential-deploys to test create/delete across deploys - Add protoLogs replacement to stabilize flaky telemetry timing - Regenerate out.requests.txt golden files Co-authored-by: Isaac

If CreateDeployment fails, the workspace should not contain a dangling deployment ID pointing to a non-existent server record. Co-authored-by: Isaac

The old lock.Acquire mutator checked for fs.ErrPermission and fs.ErrNotExist and reported possible permission denied errors. This was lost when refactoring to the DeploymentLock interface. Co-authored-by: Isaac

Add print_requests.py cleanup at the end of each script to clear remaining recorded requests, preventing out.requests.txt from being generated as a golden file. DMS requests are already printed inline in output.txt. Co-authored-by: Isaac

Open out.requests.txt with explicit utf-8 encoding to handle non-ASCII characters in request bodies. Co-authored-by: Isaac

Regenerated with Python 3.11 after fixing the UnicodeDecodeError. The output.txt files now contain the inline DMS request assertions without the Python traceback errors. Co-authored-by: Isaac

Co-authored-by: Isaac

shreyas-goenka temporarily deployed to test-trigger-is March 26, 2026 16:52 — with GitHub Actions Inactive

shreyas-goenka temporarily deployed to test-trigger-is March 30, 2026 20:00 — with GitHub Actions Inactive

shreyas-goenka had a problem deploying to test-trigger-is March 31, 2026 21:18 — with GitHub Actions Failure

shreyas-goenka had a problem deploying to test-trigger-is March 31, 2026 21:22 — with GitHub Actions Failure

shreyas-goenka had a problem deploying to test-trigger-is March 31, 2026 21:45 — with GitHub Actions Failure

shreyas-goenka had a problem deploying to test-trigger-is March 31, 2026 22:25 — with GitHub Actions Failure

shreyas-goenka had a problem deploying to test-trigger-is March 31, 2026 23:12 — with GitHub Actions Failure

shreyas-goenka had a problem deploying to test-trigger-is April 1, 2026 00:25 — with GitHub Actions Failure

shreyas-goenka had a problem deploying to test-trigger-is April 1, 2026 09:18 — with GitHub Actions Failure

shreyas-goenka commented Apr 8, 2026

View reviewed changes

shreyas-goenka commented Apr 10, 2026

View reviewed changes

shreyas-goenka force-pushed the shreyas-goenka/deployment-metadata-service branch 11 times, most recently from 4bbbe9c to 7b26260 Compare April 14, 2026 21:15

shreyas-goenka commented Apr 14, 2026

View reviewed changes

shreyas-goenka marked this pull request as ready for review April 15, 2026 00:08

shreyas-goenka requested review from andrewnester and pietern and removed request for andrewnester and pietern April 15, 2026 00:09

shreyas-goenka requested a review from denik April 15, 2026 00:09

shreyas-goenka requested a review from andrewnester April 15, 2026 00:09

shreyas-goenka added 15 commits April 15, 2026 00:09

Restore comments and whitespace removed unnecessarily from deploy/des…

e7d2877

…troy phases Co-authored-by: Isaac

Move LoadStateFromDMS from lock package to statemgmt package

1f936e6

LoadStateFromDMS is a state-loading function, not a lock function. Moving it to statemgmt where it belongs alongside other state management code. Co-authored-by: Isaac

Skip GetDeployment call for fresh deployments

6e19c93

When we just created the deployment, LastVersionID is necessarily empty so we can start at version "1" directly. Co-authored-by: Isaac

Return error for bind and unbind with deployment metadata service

0386ee6

Co-authored-by: Isaac

Write deployment ID to workspace only after CreateDeployment succeeds

e85a6e0

If CreateDeployment fails, the workspace should not contain a dangling deployment ID pointing to a non-existent server record. Co-authored-by: Isaac

Restore permission error handling in workspace filesystem lock

79b4027

The old lock.Acquire mutator checked for fs.ErrPermission and fs.ErrNotExist and reported possible permission denied errors. This was lost when refactoring to the DeploymentLock interface. Co-authored-by: Isaac

Fix UnicodeDecodeError in print_requests.py on systems with ASCII locale

5e7c839

Open out.requests.txt with explicit utf-8 encoding to handle non-ASCII characters in request bodies. Co-authored-by: Isaac

Regenerate DMS acceptance test output files

b4dc993

Regenerated with Python 3.11 after fixing the UnicodeDecodeError. The output.txt files now contain the inline DMS request assertions without the Python traceback errors. Co-authored-by: Isaac

Print DMS requests after each deploy in sequential-deploys test

ffefb37

Co-authored-by: Isaac

Remove redundant main dms test, covered by add-resources

379ff1b

Co-authored-by: Isaac

Remove unnecessary protoLogs replacement from DMS test config

65c2aae

Co-authored-by: Isaac

shreyas-goenka force-pushed the shreyas-goenka/deployment-metadata-service branch from 7339f0b to 65c2aae Compare April 15, 2026 00:10

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Integrate deployment metadata service for locking and state#4856

Integrate deployment metadata service for locking and state#4856
shreyas-goenka wants to merge 15 commits intomainfrom
shreyas-goenka/deployment-metadata-service

shreyas-goenka commented Mar 26, 2026 •

edited

Loading

Uh oh!

eng-dev-ecosystem-bot commented Mar 26, 2026 •

edited

Loading

Uh oh!

shreyas-goenka Apr 8, 2026

Uh oh!

shreyas-goenka Apr 10, 2026

Uh oh!

shreyas-goenka Apr 14, 2026

Uh oh!

github-actions bot commented Apr 15, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

shreyas-goenka commented Mar 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Key implementation details

Test plan

Uh oh!

eng-dev-ecosystem-bot commented Mar 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

shreyas-goenka Apr 8, 2026

Choose a reason for hiding this comment

Uh oh!

shreyas-goenka Apr 10, 2026

Choose a reason for hiding this comment

Uh oh!

shreyas-goenka Apr 14, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions bot commented Apr 15, 2026

Approval status: pending

/acceptance/bundle/ - needs approval

/bundle/ - needs approval

/cmd/bundle/ - needs approval

General files (require maintainer)

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

shreyas-goenka commented Mar 26, 2026 •

edited

Loading

eng-dev-ecosystem-bot commented Mar 26, 2026 •

edited

Loading

`/acceptance/bundle/` - needs approval

`/bundle/` - needs approval

`/cmd/bundle/` - needs approval