Since Centos6.8 has moved to the vault I have been having a problem installing it with cobbler and the Updates repo (from http://vault.centos.org/6.8/updates/x86_64/) included. A few RPM into the anaconda package installation, it bombs off with "A fatal error occurred when installing the <rpm.hdr object at 0x7fcacb31b1f0> package" followed by "/mnt/sysimage/var/lib/rpm: No such file or directory". It's installing whatever package comes right after iptables, or maybe itables itself.
/mnt/sysimage/var/lib/rpm looks ok on the partially installed system. All rpm signatures in the distro and the Updates repo are good as far as "rpm -K" knows. Rebuilding the repo with createrepo doesn't help. The only workaround is to not use the Updates repo. I can reproduce this problem installing from both of my cobbler servers.
Anyone seen this? It this issue a problem with one of the Berkeley DB files in /var/lib/rpm?
Thanks, wsanders
Is the CentOS 6.8 updates repo in the Vault OK?
Is the CentOS 6.8 updates repo in the Vault OK?
- Attachments
-
- Screen Shot 2017-06-20 at 12.32.19.png (61.72 KiB) Viewed 1866 times
-
- Screen Shot 2017-06-20 at 11.29.55.png (44.78 KiB) Viewed 1868 times
-
- Posts: 10642
- Joined: 2005/08/05 15:19:54
- Location: Northern Illinois, USA
Re: Is the CentOS 6.8 updates repo in the Vault OK?
The current version is 6.9, not 6.8.
You should not hardcode the minor release in the update repo.
You should not hardcode the minor release in the update repo.
Re: Is the CentOS 6.8 updates repo in the Vault OK?
6.8 was moved to vault because it's out of date and superceded. Maybe that should be a hint...
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke
Re: Is the CentOS 6.8 updates repo in the Vault OK?
Well, I confirm that the repo itself is more or less OK, since I can upgrade a system from the updates repo after installation without the upgrade repo.
Sooner or later you are going to work with lazy or superstitious developers that won't allow you to patch your legacy OSes. This setup, using the updates in the vault as a separate repo when kickstarting, works fine in Centos 6.2, which I still have to support as well, and with Centos 6.8 as of about three months ago.
Long ago there was a procedure to merge updates and base RPMs into a single distro. I'll look for my notes on that. I know that if you just copy over the update rpms on top of base and run createrepo it will nuke the groups.xml file and I don't know how to recreate that in the right place. Does anyone still remember how to do that?
Sooner or later you are going to work with lazy or superstitious developers that won't allow you to patch your legacy OSes. This setup, using the updates in the vault as a separate repo when kickstarting, works fine in Centos 6.2, which I still have to support as well, and with Centos 6.8 as of about three months ago.
Long ago there was a procedure to merge updates and base RPMs into a single distro. I'll look for my notes on that. I know that if you just copy over the update rpms on top of base and run createrepo it will nuke the groups.xml file and I don't know how to recreate that in the right place. Does anyone still remember how to do that?