Fix the issue with openldap pkg install on ssh and http proxy VMs

Issue:

Both ssh and http proxy deploys in staging cluster failed with one of packages can't be installed:

https://gitlab.rc.uab.edu/rc/hpc-factory/-/blob/main/ansible/roles/ldap_config/tasks/main.yml?ref_type=heads#L10

[ 48.571990] cloud-init[997]: failed: [127.0.0.1] (item=openldap-servers) => {"ansible_loop_var": "item", "changed": false, "failures": [], "item": "openldap-servers", "msg": "Depsolve Error occured: \n Problem: conflicting requests\n - nothing provides openldap(x86-64) = 2.6.6 needed by openldap-servers-2.6.6-3.el9.x86_64 from epel", "rc": 1, "results": []}

Cause:

The openldap package has been updated to version 2.6.8 in the epel. However, openldap-servers pkg is not updated in epel yet. Since the openldap(x86-64) = 2.6.6 is removed due to the update and openldap-servers-2.6.6 depends on openldap-2.6.6, we can't install openldap-servers pkg.

Proposed solution:

1. Either remove the openldap-servers pkg from the list since we only use openldap as a client, and we are not setting up an openldap-server (already provided by Bright on the headnode) OR

2. Build the pkg yourself and upload it to S3. Update the pkg list in ansible task to install the pkg from the s3 location.

added Sprint 25-11 label

changed the description

I built the openldap-servers pkg and uploaded it to S3 in case if we decide to still keep the openldap-servers pkg to avoid any lookup failures in the event of pkg dependencies for openldap on openldap-servers. It is accessible here http://s3.lts.rc.uab.edu/openldap-servers-2.6.8/openldap-servers-2.6.8-1.el9.x86_64.rpm

Note: I still feel like we can safely remove the openldap-servers pkg. However, this hasn't been tested yet.

mentioned in merge request !199 (merged)

We decided to remove the openldap-servers pkg install because we only use openldap client for lookups

mentioned in commit 44e0efb4

closed with merge request !199 (merged)