…n the Administrator/superadmin role
now-superfluous install step from wiki.
A recent commit updated all resources to be uri encoded (930851c). Unfortunately, distribution names were already being encoded. As a result, this double encoding caused problems for that case. This commit removes the encoding from pulp requests for distributions.
It reproduces an issue with space encoding.
…when force upload is already set
In Katello, it is possible to have multiple repositories with the same name (e.g. repoX) across multiple products. This commit contains changes to improve Katello's ability to support that with System Templates. Essentially, allow the user to pick a specific repo and add it to a template. This applies when selecting the repo from the left content tree or when using the auto-complete search box from within the right template tree. In the case of auto-complete, as the user types the repo name, Katello will provide the repos that match along with the product they are part of. If the user picks from the auto-complete results, only the selected repo will be added to the template. If the user ignores the auto-complete results and adds the repo explicitly based on what they typed, then Katello will add all repos that match to the template.
- fixed wrong strings created by copy-paste from other commands - it is now possible to update only date and time together
It's causing problems when wiping the database. The fix is comming in cp-0.5.16-1. Enabling this test again after new CP release.
Only marketing products must deal with deleting product - content assignment in CP. For engineering products there are repositories created, which delete the content when they are removed. This change prevents trying to delete the content twice while deleting product. The other posibility would be to prevent deleting of conent in repository in case product is being deleted and let product handle it all, but there would be necessary a mechanism for finding out if the repo orchestration happens during product orchestration or not - when not repo should handle content deletion itself.
When running the dealyed jobs with commend: ./script/delayed_job run the logs of delayed job worker (including Rails.logger) are written into log/development_delayed_jobs.log and sql logs go into log/development_delayed_jobs_sql.log. This helpes better track what happnes while runnin async operation, especially calling of remote services.