I noticed a number of servers subscribed to our production OLM repos were not showing updates that the test/development were showing when I subscribed to them. The repos appear to be installed on the server and yet yum -v repolist show an update date over a month ago,when I started seeing the behavior. I totally rebuilt all pf the channels and repos using the exact same data as the working server and still, the number of packages reported per channel do not match, and the yum check-update even on that one OLM VM is not showing any updates. I have tried yum clean all, deleting the cache folder and a database rebuild and still the latest packages that whos in the GUI do not show up via yum list.
broken system:
Looking for repo options for [oracle7]
Repo 'oracle7' setting option 'enabled' = '1'
Repo 'oracle7' setting option 'gpgcheck' = '1'
Repo 'oracle7' setting option 'timeout' = '120'
Repo-id : oracle7
Repo-name : oraclelinux7
Repo-updated : Fri May 31 02:47:19 2024
Repo-pkgs : 21,623
Repo-size : 48 G
Repo-baseurl : http://10.10.16.67/XMLRPC/GET-REQ/oracle7
Repo-expire : 21,600 second(s) (last: Fri Jun 28 13:47:41 2024)
Filter : read-only:present
[root@dalvolm01 yum.repos.d]# yum list available|grep selinux-policy-dev
selinux-policy-devel.noarch 3.13.1-268.0.19.el7_9.2 oracle7
Even though the correct version is in the production server repo:
working system:
Looking for repo options for [oracle7]
Repo 'oracle7' setting option 'enabled' = '1'
Repo 'oracle7' setting option 'gpgcheck' = '1'
Repo 'oracle7' setting option 'timeout' = '120'
Repo-id : oracle7
Repo-name : oraclelinux7
Repo-updated : Fri Jun 28 04:03:53 2024
Repo-pkgs : 21,727
Repo-size : 49 G
Repo-baseurl : http://10.1.49.169/XMLRPC/GET-REQ/oracle7
Repo-expire : 21,600 second(s) (last: Fri Jun 28 12:04:47 2024)
Filter : read-only:present
Repo-exclude : jta*
[root@houvolm01 scripts]# yum list available | grep selinux-policy-dev
selinux-policy-devel.noarch 3.13.1-268.0.23.el7_9.2 oracle7
I have subscribed to the production repo for monthly patching, but that defeats the purpose of keeping traffic with the production datacenter subnet
Any ideas? Thanks in advance!