Make clusters start succeed when the cluster is already running - #6288
Draft
TangoEnSkai wants to merge 1 commit into
Draft
Make clusters start succeed when the cluster is already running#6288TangoEnSkai wants to merge 1 commit into
clusters start succeed when the cluster is already running#6288TangoEnSkai wants to merge 1 commit into
Conversation
The API rejects a start request for a cluster that is not TERMINATED with INVALID_STATE, while the command's help states that nothing happens in that case. Re-read the cluster on INVALID_STATE and report success when it is already RUNNING; every other rejection is returned unchanged. The test server accepted clusters/start in any state, so it could not express this; it now returns the same INVALID_STATE error as the API. Fixes databricks#1372 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
TangoEnSkai
force-pushed
the
clusters-start-idempotent
branch
from
August 17, 2026 13:16
f574a7a to
06ef290
Compare
Contributor
|
An authorized user can trigger integration tests manually by following the instructions below: Trigger: Inputs:
Checks will be approved automatically on success. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Changes
databricks clusters starton a cluster that is already running now reports success and prints the cluster, instead of failing withCluster <id> is in unexpected state Running.. Any otherINVALID_STATErejection (a terminating cluster, for example) is still returned unchanged.The test server now models the API's
INVALID_STATErejection ofclusters/startfor a cluster that is not terminated; it previously accepted the call in any state.Fixes #1372.
Why
The command's own help text states "If the cluster is not currently in a TERMINATED state, nothing will happen", but the API rejects the request, so a script that just wants the cluster up has to match on the error message. The direct engine already treats an already-running cluster as a no-op (
bundle/direct/dresources/cluster.go), so this aligns the standalone command with both the documented contract and bundle behavior.The check is deliberately narrow: it re-reads the cluster after
INVALID_STATEand only swallows the error when the state isRUNNING. A cluster that isPENDINGorRESTARTINGstill errors out, because succeeding there would mean waiting for a state transition rather than reporting a no-op.Tests
New acceptance test
acceptance/cmd/workspace/clusters-start, covering both an already-running cluster and a terminated one. Reverting only thecmd/workspace/clusterschange makes it fail with the exact error from the issue:./task fmt,./task checksand./task lintare clean../task testpasses locally except forbundle/templates-machinery/supported-url, which fails the same way on an unmodifiedorigin/maincheckout here because my network path answers the intentionally invalid host with a proxy error instead of a DNS failure.This PR was written by Claude Code, then reviewed and tested by me.