Testing and troubleshooting - Jira Cloud
Testing the integration
Pick a low-priority error from a test project and create a ticket from it. In Jira, check the issue type (it should be close to 'Bug', not an epic or a sub-task), the description content, and whether the assignee and starting status match the project configuration. To check the sync back, switch that ticket to a Done-category status by hand in Jira and wait up to five minutes - the error status in DockRay should flip to resolved with no action on your part.
Common problems
The ticket gets an unexpected issue type
DockRay has to recognise the 'bug' type independently of Jira's language - Jira Cloud translates type names into the account's language, and the type list starts with the epic. That is why the integration first rejects sub-tasks and everything at epic or initiative level, and only then searches what is left for a name close to 'Bug', then 'Task'. If a Jira project's issue type scheme has neither, the ticket still will not become an epic - it gets the first available type at the ordinary work level, sometimes something like 'Story'. Fix: add an issue type matching a bug to the Jira project.
The Jira project key gets rejected
The key must start with an uppercase letter and contain only uppercase letters, digits and underscores. A key typed in lowercase, or with a space, hyphen or other special character, gets rejected - that is not over-eager validation, it protects against steering the request outside the intended API endpoint.
The error status does not come back to the panel right away
That is by design. Status sync is periodic, not instant - Jira Cloud's dynamic webhooks require a registered OAuth/Connect app, which cannot be set up with just the client's API token. The thor:sync-jira command polls linked tickets every five minutes, so a delay of up to five minutes is normal, not a failure.
The connection fails with an authentication error
Check whether the e-mail address and the API token belong to the same Atlassian account, whether the token has been revoked in Atlassian's settings, and whether the instance address really is on the *.atlassian.net domain - any other address is rejected before the integration even tries to connect.
The integration stops working after a project settings change
This happens after switching a project from its own credentials to inheriting the account's, when the account-level connection has meanwhile disappeared too. Check both levels in turn: the project configuration and the account configuration - a project with neither has nowhere to take credentials from.
Still not working
Work through the troubleshooting checklist, and if that does not help, write to us. Include the project name, the integration version and roughly when you ran the test: it shortens the way to an answer.