Hide Forgot
Description of problem: When i select rerun failed hosts the job invocation is not updated with the new result. Instead a new job invocation is created for the failed hosts. From a user point of view my expectation is a 'rerun' means to re-run it in the context of the same invocation and update the results of that job invocation. With the current design to create a new invocation the overall result of the job is lost, because as an user i have now to overlay the results of multiple job invocations to see the total result. For a good and user friendly reference implementation see Atlassian Bamboo how it handles run incomplete or failed or rerun. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. Run job template on 5 hosts, make sure 3 hosts will fail 2. Fix the 3 failing hosts to make sure the job succeeds 2. Click rerun failed templates 3. Actual results: A new job invocation is started for the 3 failed hosts Expected results: The current job invocation is updated and shows 100% success Additional info:
The suggested approach doesn't fit into the overall design all that well. Job invocations and tasks are considered immutable once finished. We could provide some kind of link between the old and the rerun-ed job invocation, so the user could see how they relate. Would that work for you?
Links still do not provide the total overview of the execution on all hosts. I still loose the overall result for all servers. An implementation with job invocation versioning will work. That is similar to the contentview versions already existing in satellite In the backend each invocation is still an unique id. But from a user point of view you see a user friendly version and run number: Job #1 run #1 Job #1 run #2 Job #1 run #3 For me as a user this provides a more clean and quick approach to re-run jobs that i already did in the past and then even compare the number results for that job. The current implementation that each job is an unique run starting from a template does not provide this kind of correlation of first and subsequent re-runs.
Created redmine issue http://projects.theforeman.org/issues/18522 from this bug
Thank you for your interest in Satellite 6. We have evaluated this request, and we do not expect this to be implemented in the product in the foreseeable future. We are therefore closing this out as WONTFIX. If you have any concerns about this, please feel free to contact Rich Jerrido or Bryan Kearney. Thank you.