some time ago we implemented HTTP proxy feature which was IIRC supposed to be used by (some) Compute Resources. Can someone explain me what is the plan with this feature? Which CRs are currently used? Which are planned?
Can we do better in explaining to the users what these HTTP proxies are for on the index page?
I believe you already got the answer, this is used for proxying the communication between Foreman and CR. By looking at the code, every compute resource has the relation defined but I only see http_proxy_field helper used for form of EC2. IIRC @Justin_Sherrill worked on this and may know more details, but I think this is the only CR that supports it.
Note that there’s also a global HTTP proxy setting that can be used in order to proxy all outgoing traffic which should work with all compute resources that uses HTTP based communication. That is though for all HTTP traffic, not just compute resources. IIRC one can black list FQDNs there too.
Yes, only EC2 support was added initially, the hope was that others
would follow as well as katello and docker plugins would convert to use
it. I’d still like to move Katello over to use it (since it doesn’t use
the global setting for pulp syncs), but has its own configuration for that.
IIRC It was also started before the global proxy change was done, but
was merged afterwards (since it was a larger change).