Hi, all,
Hoping to get some assistance root causing this issue. We have a set of CentOS 7 based systems in AWS that we're using the leapp tool to migrate to AlmaLinux 8.7. Following migration, the root filesystem is ending up being mounted read-only, and I'm having
a devil of a time figuring out why. There are no helpful (to me) messages in dmesg, or on the console during boot. Root on these machines is xfs - the filesystems appear to be intact, and running "sudo mount -o rw,remount /" following boot succeeds straight
away.
I can see that initrd-switch-root.service runs during initialization (I see "Reached target Switch Root.", followed by "Starting Switch Root..." on the console, and then after, can "systemctl status initrd-switch-root.service", and it indicates that the systemctl
command in ExecStart completed w/ a 0 rc).
The AlmaLinux 8.7 distro is using systemd 239 - and CentOS 7 that was migrated from was/is on systemd 219.
Also, for reference, we have another set of systems that we're performing the same migration on, but they're configured a bit differently (ext4 root partition, for starters) - those come up just fine with / mounted rw after reboot. And, starting with a bare
minimal C7 system in a local VM, but with filesystems configured the same as our problem children (xfs root, single partition), it also migrates just fine. So, there's something off with the problem children, but I'm having trouble nailing it down. Understanding
why root wants to be ro on those systems would hopefully point me at the culprit.
Thanks for any help!
Dave
|