QA

User story¶

In order to test non-trivial changes on maintenance job groups we could previously use the openQABot to run a custom metadata definition against our own openQA development instances. This was very useful, because it reveals issues soon and one has not to wait for the next scheduling round to test the changes (and run into trouble because something broke).

This however is not possible anymore with the new QA bot. We need this feature, otherwise the implementation of new scenarios is slowed down drastically, as changes can not be tested outside of the production environment (OSD) and can only happen at the typical scheduling tasks (i.e. 1 scheduling round per day).

Acceptance criteria¶

• Add the ability to bot-ng to schedule test runs on a custom openQA instance (e.g. configurable with the --remote) parameter
• Remove the necessity of the QAM Dashboard token for such runs to prevent users from accidentally updating the dashboard with non-production data

Further details¶

I've already drafted a basic implementation in a merge request, Ondrej however mentioned that this will not work due to editiong conflicts with the QEM Dashboard