Pipeline stuck on "Wait for Manifest to Stabilize" Task
Issue
When executing a pipeline the Wait For Manifest To Stabilize
stage gets stuck until the timeout is reached.
Cause
This can be caused by a few different reasons. In order to narrow down the potential root cause, please check the following:
- If the execution was initiated by non-admin user and when an admin user starts the the pipeline, it gets stuck
- If manually triggering the pipeline shows the same behavior
- If there is a difference in behavior when checking the above, then the issue may be due to the moniker naming convention.
What happens is that the request to fetch a manifest from Clouddriver does two permission checks:
- Does the user have permission to the account in question?Does the user have permission to the app specified on the manifest's app
- In cases where a
moniker.app
is explicitly specified on the manifest, then this is the app that is used.* In cases where there is nomoniker.app
specified, then the app is determined by parsing usingfrigga
as per this article.