You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
but this is not idempotent and this seems to do lots of work on each run. It would look better to allow using the purge parameter for configuring repositories exclusively. Thanks.
The text was updated successfully, but these errors were encountered:
but this is not idempotent and this seems to do lots of work on each run
This is correct at the moment, yes.
It would look better to allow using the purge parameter for configuring repositories exclusively.
The purge attribute as available in the rhsm_repository module is a complete mess:
its semantics were bolted on top of the existing enable/disable logic, and the resulting flow is overly complicated
its semantics are not even totally clear: some of the combination of parameters with it create unintuitive results
I did some changes a couple of months ago to that module, and the changes were from easy, even because of purge
So, with both my rhc role maintainer hat, and my rhsm_repository upstream module maintainer hat, I disagree that purge is a good solution for this issue. The fact that the role allows users to specify the repositories this way gives some room for potential optimization, either in the role itself or in rhsm_repository; this will need some design/thoughts.
It is not possible to configure repositories exclusively in an idempotent fashion today with rhc. This works:
but this is not idempotent and this seems to do lots of work on each run. It would look better to allow using the purge parameter for configuring repositories exclusively. Thanks.
The text was updated successfully, but these errors were encountered: