loading repo 'rhel-8-for-x86_64-baseos-rpms' failure: loading of MD_TYPE_PRIMARY has failed

Latest response

Hello everyone,

I'm doing test for implementing RHEL 8 in my server park.
I'm having a weird probleme, the rhel-8-for-x86_64-baseos-rpms have stopped working and when i do whatever command with yum or dnf i got this message :

[root@sdtestrhel8 ~]# yum update -v
Loaded plugins: builddep, changelog, config-manager, copr, debug, debuginfo-install, download, generate_completion_cache, needs-restarting, playground, product-id, repoclosure, repodiff, repograph, repomanage, reposync, subscription-manager, uploadprofile
Updating Subscription Management repositories.
YUM version: 4.2.17
cachedir: /var/cache/dnf
User-Agent: constructed: 'libdnf (Red Hat Enterprise Linux 8.2; generic; Linux.x86_64)'
Red Hat Enterprise Linux 8 for x86_64 - BaseOS (RPMs)                                                                                                                                                         28 kB/s | 1.5 kB     00:00
reviving: 'rhel-8-for-x86_64-baseos-rpms' can be revived - repomd matches.
**loading repo 'rhel-8-for-x86_64-baseos-rpms' failure: loading of MD_TYPE_PRIMARY has failed.**
Error: Loading repository 'rhel-8-for-x86_64-baseos-rpms' has failed

I've tried disabling zchunk for yum but no luck on this side.

It happend like 10-15 minutes after using https://access.redhat.com/solutions/4832211 solution for EPEL 8 repo not working, but for me it seem like it's not related because BaseOS repo worked during a moment after the resolution for EPEL 8.



Hello Nathan,

That solution you mentioned above (#4832211), are you using a Red Hat Satellite server? Additionally, (just curious here), it seems this issue occurred when you attempted to make EPEL available for your system? Is that correct? If so, what happens if you temporarily disable the EPEL repository? Does this go away?

Nathan, is your other post at https://access.redhat.com/discussions/5304851 somehow related to this?

Initially, I'd recommend submitting a case to Red Hat for this issue after ruling out if (yes or no) temporarily removing the EPEL repository causes the issue to go away.

Nathan, please let us know how this goes, or if I've missed something here in my reply.

Kind Regards,

Hello Hinton,

Effectively I forgot to put some informations here : - I'm using a satellite server (6.6) - I don't thinking this is related to the EPEL (didn't change anything to disable the EPEL repo, on client side and on the satellite)

I've also tried to resync the repo with hammer repository synchronize --validate-contents=true --id=ID_OF_BASEOS_REPO but no luck here too.

Additionnaly to this I just saw that the /var partition of the satellite is fully used to 100% , I'm fixing this to see if it's related (hope this is) and will post how it goes.

Me and my team where thinking of opening a case, but we would like to see if we can fix it ourselves before calling Red Hat

Thank you

Hi Nathan,

After resolving the 100% full condition, run systemctl is-system-running to see if there are any failed services. I can provide some more relevant information later today.

EDITED: After resolving the 100% full condition, consider running katello-service status and if necessary katello-service restart or simply reboot the satellite if that is an option. That command restarts all satellite-related services, However, I'm not finding "katello-service" in recent documentation for satellite but see this RH solution on that command


Hi Hinton,

Sorry for not responding yesterday, the day was a bit busy and I had a lot to do.

Fixing the full /var didn't fix the repo probleme, and the systemctl is-system-running and katello-service status showed me a running and All services are running [OK] (very usefull command by the way, i wrote them on my notepad ;) )

Anyway I tryed a reboot of the satellite, but no luck here too.

Any new knowledge about this problem will be posted here.

Thank you very much for your support and your precious advises Hinton.

Just touched a thing on this :

BaseOS RPM is in a content view which is in a composite content view activated by a key marked as for 8 servers (not for 8.0 servers only, but for all 8.X versions) and my test server is a 8.2.

I made a content view for 8.2 which activate with a key for 8.2 servers only and BaseOS work just as expected...

I would like to make a key that work for all RHEL 8 servers (8.0 8.1 and 8.2) but it seems that I might missed something making it...

Will post it if ever i acheve it, but will need more tests

Hi Nathan, (I just go by RJ),

Regarding your last post on your need to make a content view. Are you familiar with the process to make a content view? Have you made a content view before in Red Hat Satellite? I'm in the process of rebuilding my satellite servers to something more current, so what my experience is, and depending on if you are using a more current version of RH Satellite, the method may differ slightly.

Now I have made activation keys that work well, but I've not done so for RHEL 8. The recipe involves assigning the proper subscriptions for the activation key for, and the repos in question. And the differences between your satellite version and mine may make this slightly different.

When I face this kind of thing, I attempt to dig through the documentation - and sometimes (if you have access to it), the videos, etc within the Red Hat Learning subscription are generally quite worthwhile on this. Sometimes, I've found the Learning Subscription methods for satellite are often .... better.

Let us know how this goes Nathan,

RJ (just "RJ")


Satellite is a supported product - perhaps consider submitting a case for what doesn't work so you will get something more efficient. The folks at Red Hat will generally always ask for a sosreport to be attached to the case.



I'm sorry I took so much time to respond.

All this was a side effect of the baseos contain view having at the same time the repo for RHEL8 servers (that fit for all 8.X) and each individuals repos for 8.0 8.1 8.2

I made a content view with only BaseOS for 8 servers and it woked just as expected !

Thank you very much RJ ;) for all you advises Regards