Skip to main content
Deployment·HTTP 500·API, FE, CLI

Deployment didn't complete

DEPLOYMENT_FAILED

The deployment finished with errors. The previous version is still serving traffic.

What this means​

The new version didn't pass health checks within the configured window, or the platform couldn't pull the container image. Spec details are in the deployment's events.

How to fix it​

Check the deployment events for the specific failure (csph deployment events <id>), fix the issue, then redeploy. Common causes: failing health checks, missing environment variables, bad image tag.

Retry behavior​

Fix the underlying issue first. Redeploying without changes will hit the same error.

How it appears in the response​

ComputeSphere returns errors as RFC 7807 Problem Details. The type field is the docs URL for this page — switch on code in your client.

{
  "type":    "https://docs.computesphere.com/errors/DEPLOYMENT_FAILED",
  "title":   "Deployment didn't complete",
  "status":  500,
  "code":    "DEPLOYMENT_FAILED",
  "detail":  "The deployment finished with errors. The previous version is still serving traffic.",
  "instance": "/v2/…",
  "request_id": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
  "trace_id":   null
}

The Content-Type header is application/problem+json. On field-level validation errors a non-empty errors[] array also appears, each entry carrying its own docs_url for the offending field.