Composite Content View vs Content View


I’m trying to wrap my head around the use of a composite content view versus a content view.

If we have the following repos:

repo a (Base rhel7 os)
repo b (epel)
repo c (custom)

What is the benefit of adding all 3 to a single content view versus creating a separate content view for each repo (so 3 content views) and then adding them all to a composite content view?


Separate CV + Composite CV => publish a new version of e.g. repo a without touching repo b and repo c
Single CV => easier to handle.

Have a look at Feature #21994: Auto publish Composite if component updates - Katello - Foreman which is available since Katello 3.7.0

1 Like

It’s really up to you - thankfully it’s relatively easy to experiment with both.

I have a relatively conservative internal business, so some parts - the fragile parts - we put into separate CVs so that we could update the OS without breaking that one tool. In our case it was a particular COPR sourced version of SSSD that worked and was a couple of versions ahead of EPEL. Authentication needed to be solid and dependable and easily rolled back. So we made it separate.

But MariaDB, El Repo, Docker, Zabbix are all in a CCV called “Extras” because at no point are they mission critical.

1 Like

Thank you both – that helps.