Update repolist for use with micrdnf in offline satellite env

[budgester]

How do you go about updating the repos that microdnf uses when your in a restricted environment with ansible-builder. We currently use satelite for our repos.

/usr/bin/microdnf install -y --nodocs --setopt install_weak_deps=0 gcc java-17-openjdk libcurl-devel libssh-devel libxml2-devel make openssl openssl-devel python38-Cython python38-devel unzip
--
Downloading metadata...
error: cannot update repo 'ubi-8-baseos-rpms': Cannot download repomd.xml: Cannot download repodata/repomd.xml: All mirrors were tried; Last error: Curl error (6): Couldn't resolve host name for https://cdn-ubi.redhat.com/content/public/ubi/dist/ubi8/8/x86_64/baseos/os/repodata/repomd.xml [Could not resolve host: cdn-ubi.redhat.com]
Error: error building at STEP "RUN /output/scripts/assemble": error while running runtime: exit status 1

[Akasurde]

@budgester Thanks for reporting this issue. Can you access the repomd.xml from outside (the container)? Also, do you have any proxy in place? Please provide the ee.yml and the image used for this issue? Thanks in advance.

[budgester]

We don't use a proxy, this is an internet restricted environment. So we have an jfrog artifactory for python packages and redhat satellite for yum packages.

If I was using dnf I'd have more tools but the ee image only has microdnf baked into it. So I need to add a satelite repo and certs to enable microdnf to have the ability to install yum packages.

I can access the satellite repomd.xml from the host machine, but not within the container.

---
version: 3

dependencies:
  galaxy: requirements.yml
  python: requirements.txt
  system: bindep.txt

images:
  base_image:
    name: artifactory.mydomain.co.uk/redhatcontainers/ansible-automation-platform-23/ee-supported-rhel8:latest

options:
  package_manager_path: /usr/bin/microdnf

additional_build_steps:
  prepend_base:
    - RUN $PYCMD -m pip config debug
    - RUN cat /etc/yum.repos.d/redhat.repo
    - COPY pip.conf /etc/pip.conf
    - COPY redhat.repo /etc/yum.repos.d/redhat.repo
    - RUN $PKGMGR makecache --disablerepo=* --enablerepo=rhel-8-for-x86_64-baseos-rpms
    - RUN $PKGMGR repolist
    - RUN $PYCMD -m pip config list

[budgester]

This seems to have helped..

#350 (comment)

I think I'm down to environmental issues now.

Installing: openssl-devel;1:1.1.1k-9.el8_7;x86_64;rhel-8-for-x86_64-baseos-rpms
Installing: libxml2-devel;2.9.7-16.el8;x86_64;rhel-8-for-x86_64-appstream-rpms
error: SQLite error on "/var/lib/dnf/history.sqlite": Reading a row failed: database or disk is full
Error: error building at STEP "RUN /output/scripts/assemble": error while running runtime: exit status 1
An error occurred (rc=1), see output line(s) above for details.

[Akasurde]

@budgester Should we close this issue as an environment issue?

[budgester]

I think it should have a documentation update. I'll try and get a PR to update the docs together.

[Akasurde]

@budgester Any updates? Thanks.