pulp_rpm icon indicating copy to clipboard operation
pulp_rpm copied to clipboard

Better documentation for simple mirroring/caching of upstream repos

Open abctaylor opened this issue 2 years ago • 5 comments

Hi, this might be specific to pulp overall, than just the operator.

Could it be documented better (I need help here myself too) how to simply mirror/front some upstream repo? for example, take the Almalinux baseos repo. It has a baseurl (http://almalinux.mirror.katapult.io/8.9/BaseOS/x86_64/os/) and mirrors (https://mirrors.almalinux.org/mirrorlist/8/baseos).

What precisely must be done to provide this on prem via Pulp?

abctaylor avatar Apr 02 '24 16:04 abctaylor

I would recommend filing issues in the pulpcore or pulp_rpm. This issue tracker is specifically for pulp-operator.

Here is a link to the new documentation site that we are working on:

https://staging-docs.pulpproject.org/pulp_rpm/docs/user/tutorials/01-create_sync_publish/

dkliban avatar Apr 05 '24 14:04 dkliban

I am going to transfer this issue to pulpcore.

dkliban avatar Apr 05 '24 14:04 dkliban

Note: belongs to RPM since the "interesting" flags/howto are pulp_rpm-specific.

ggainey avatar Apr 18 '24 15:04 ggainey

It looks like most of the relevant information is already in our docs page, so I suppose the action item here is to simplify and make it flow better so that this information is more obvious.

@abctaylor @dkliban On that page, do you feel anything in particular is missing or unclear? We definitely want to make sure they're as complete and clear as possible.

dralley avatar Apr 18 '24 15:04 dralley

do you feel anything in particular is missing or unclear?

I'm evaluating your project as one option for hosting a private pull-through RPM cache. One thing I'm not seeing is how to store the artifacts in S3.

In fact, from the docs I wouldn't know your project supports S3 at all, except for one small hint and a dead link on this page ("locally or on S3").

I personally am hoping for a single doc that, end to end, spells out how to create such a S3-backed cache. Ideally with numbered steps and small explanations as to how each step enables the next.

I'll be targeting running pulp in an EKS cluster. As such the tutorial I'm looking for starts with an EKS cluster with no running pulp installation, and ends with a working RPM pull-through cache. I could probably cope with a different deploy target, but that is the scope I'm hoping for.

phs avatar Jul 11 '24 18:07 phs