2025-01-08 Rename o*f-new#
When we installed the new instances for obf/opf/opff, we named the new instances with a -new name.
This causes problem for certain systemd services uses the hostname to get the environment files. The trick was to create a folder with the new name and a symlink to the configuration file. This remains a quirk in current installation.
So to do some cleanup, we will:
- rename the old instances with a -old suffix
- rename the new instances to remove the -new suffix
Renaming old instances#
We are speaking of containers 110, 111, 112
First, using proxmox interface, I saw in options that those old containers are still set to start at boot. So I unset the option.
Then I go to DNS and change the hostname to add a -old
suffix.
Renaming new instances#
We are speaking of containers 116, 117, 118
I still use the proxmox interface, but I do one by one:
- rename the container to remove the
-new
suffix - reboot the container
- verify the website still function (after some time)
- use
systemctl status
to check everything is running (along withsudo systemctl list-units --state=failed
) - remove the /o*f-new directory in the container
Annex making sytemd status running#
While at it, looking at systemctl status, the following services are failed:
- sys-kernel-config.mount
- systemd-journald-audit.socket
and it's perfectly normal because we didn't gave capabilities for that to LXC
(would need capabilities
sys_rawio
andaudit_read
) Thanks to this blog post, I infer that he best way to deal with it is not to disable thoses services, but to drop capabilities for the container.