Migrate from RHEL7 to RHEL8 TEST system

[kketchmark]

I have built a new system with RHEL8 installed 5.11.2 (since this is what my old system was, but after a dnf update it updated Nagiosxi to 5.11.3. Nagios look fine so I did a backup of PROD and copy'd the backup file over to TEST system to run the restore(restore_xi-sh 1699286699.tar.gz). That went good other then saying to download and run the restore_repair.sh script since it was a RHEL7 backup and restore server was RHEL8. I did run the restore_repair.sh and everything went fine. But my issue is that the version is saying 5.11.2 and when I try to upgrade it says Nothing to do.

If I run a dnf list installed |grep nagios ~ it comes back with this. Looks correct to me.
nagios-repo.noarch 8-1.el8 @System
nagiosxi.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-mrtg.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-nagioscore.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-nagiosplugins.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-nagvis.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-ndoutils.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-nrds.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-nrpe.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-nsca.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-nxti.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-pnp.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-shellinabox.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-wkhtmltox.x86_64 5.11.3-1.el8 @nagios-base
nagiosxi-wmic.x86_64 5.11.3-1.el8 @nagios-base

Could it just be that the file /usr/local/nagiosxi/var/xiversion is not updated to 5.11.3?
I see the /usr/local/nagiosxi/var/xiversion has the same time stamp is what is on PROD so I assume the restore pulls that file with the backup/restore.

Does anyone know what I should do?

[swolf]

Hi @kketchmark, thanks for reaching out.

I haven't seen an issue exactly like this before, so I'm speculating, but I would guess that xiversion and in fact all other files were on 5.11.3 at some point, and then were downgraded to 5.11.2 when you restored your 5.11.2 backup. However, that restore process doesn't change your RPM packages, so yum/dnf "thinks" you're using package version 5.11.3. The "Nothing to do" message comes from yum, and is only checking the RPM package version.

My recommendation to you is that you either roll your test system back to when 5.11.2 was installed or build a new one and prevent any packages from our repository from updating beyond 5.11.2, then run the migration again with your 5.11.2 backup. Alternatively, if your RHEL7 system is still operational, you could update your RHEL7 system to 5.11.3 and restore from that backup instead. Basically, you need the version of your backup and the installed version to match.

I hope this helped - please let me know if anything is unclear or if you have any further questions or concerns.

[kketchmark]

I tried to downgrade it without any luck. I will destroy this system and build a new one and see what happens.

[kketchmark]

Complete rebuilt at version 5.11.2 with import of data without issues, upgraded to 5.11.3 also without issues.

[jsimon]

Hi @kketchmark,

It sounds like you have resolved your issue but I wanted to check in and make sure you're good to go before locking the topic.