Spinnaker "Stage ... failed" pipeline error in CI
Spinnaker marked a stage as TERMINAL because one of its tasks reported failure. The pipeline halts at that stage and the details tab holds the underlying task exception that actually stopped it.
What this error means
A CI-triggered Spinnaker pipeline stops with "Stage 'Deploy' failed" (or the stage name you gave it) and the execution shows a red TERMINAL status. The stage details expose the failing task.
Exception ( Deploy )
Stage Deploy in pipeline deploy-prod failed.
Status: TERMINALCommon causes
A task inside the stage reported failure
Stages are made of tasks (bake, deploy, monitor). If any task returns TERMINAL, Orca marks the whole stage failed and stops the pipeline.
The stage was configured to halt the pipeline on failure
With "halt the entire pipeline" set on failure, a single stage error stops everything rather than continuing or branching.
How to fix it
Open the stage details and read the failing task
- Open the execution in Deck and expand the red stage.
- Read the "Exception" block in the stage details for the real task error.
- Fix the underlying task (bake, deploy, or monitor) and re-trigger from CI.
Choose an explicit failure strategy
Set the stage "If stage fails" option deliberately so a known-flaky stage does not stop the whole pipeline unexpectedly.
# In the stage config, under "Execution Options":
# If stage fails: [halt the entire pipeline | halt this branch | ignore the failure]How to prevent it
- Set a deliberate failure strategy on every stage.
- Surface the failing task exception in your CI logs, not just "Stage failed".
- Add a Manual Judgment before irreversible stages so a bad deploy pauses.