nagiosxi 2024R1 and mod gearman

[elinagios]

Hello

I have currently nagiosxi 5.11.2 version with Centos 7. The setup also includes mod-gearman with and outside gearman proxy, i know, i know, mod gearman is deprecated and should not be used. Im setting up a migration to new OS, as centos 7 support will end soon ,where there is no mod-gearman. My question is, can i upgrade nagiosxi to 2024R1 version before that or the mod-gearman will certainly brake with new version and i need to migrate to new OS with no mod-gearman before?

[sgardil]

Hey @elinagios

After looking into other threads it seems like in the past upgrading to newer versions could cause some issues if they werent compatible however I don't have an instance to test this for you. The safe bet would be to migrate to a different OS at 5.11.2 and then upgrade. If you wanted you could always take a snapshot and then try the upgrade after and see what happens, that way if it doesnt work for 2024R1 then you can always just revert and run the upgrade on the new OS that doesnt have mod-gearman.

[cnorell]

As an extra guard rail and litmus test, I would disable the NEB module for Mod Gearman and ensure your XI instance is still monitoring properly before the migration, if possible. Regardless of your approach, I would take a snapshot/backup of your XI server before attempting any potentially breaking changes.

[elinagios]

I will try with migrating first and doing the upgrade later after no mod-gearman. Few questions for users.
1) Have you had good luck migrating from centos 7 to some ubuntu LTS version ? There are guides and how-to's but will you have a functional nagiosxi instance after that? No issues after running restore_repair script?
2) I did quick first migration test from centos 7 to ubuntu 22.04 and after restoring from the backup and starting the restore_repair script, it recovers rather quickly but all the service and host checks will activate and start spamming users. Now after the restore there is no web UI yet available and as the repair takes some time, the amount of alerts sent out, is too great, need a workaround. I can stop postfix, the emails wount reach the users but i also have some event handlers set in place in case of alarms, they will be sent to some outside system and at this moment they will flood those with false positives. So any way to stop the alarms and service/checks before the restore_repair script that will stay disabled after running the script? Rememer, no UI yet available before repair script, to disable alarms and service/host checks.

[swolf]

Hi @elinagios, thanks for reaching out.

With regard to Mod Gearman - I don't think we do anything that would cause issues with Mod Gearman in Nagios XI 2024. If that NEB module worked for you before, it should continue working now.

As far as the backup/restore process goes, it usually seems to be okay for going across versions of distributions (i.e. CentOS 7 -> CentOS Stream 9 / RHEL 9), but I don't personally think it's super stable for going across distributions (CentOS -> Ubuntu) - we're aware of and actively working on this. If you're already ditching Mod Gearman and don't want to use RHEL, you might also have success migrating to Oracle Linux. Regardless of the path you choose, I would highly recommend opening a case with the support team proper.

To answer your last question, if you really just need a one-off workaround, I would probably just switch out the event handler scripts themselves until the alarms die down. Most customers have something like four notification commands in total, though your system is a bit more customized. Slack has a pretty straightforward webhook system, and I'm sure other messaging clients have similar capabilities, if you're okay with dumping all of your alerts into an empty-ish room for a few minutes.

Hopefully that helps - feel free to reach out with any additional questions or concerns
-Sebastian Wolf

[elinagios]

Hi

1) OK, will plan a centos 7 to rhel 9 upgrade then. Do i still need to run the restore_repair script after restoring nagios?
What do you mean by opening a case? To check that configs are ok for migration?
2) I can not change one of the event handler types, which are sending snmp traps based on service alarms. Second might be doable, second type of handlers are shell scripts. Need web UI to do change those, so any ideas are welcome how to disable service and host checks and also alarms.

[swolf]

Hi

1) OK, will plan a centos 7 to rhel 9 upgrade then. Do i still need to run the restore_repair script after restoring nagios?
What do you mean by opening a case? To check that configs are ok for migration?
2) I can not change one of the event handler types, which are sending snmp traps based on service alarms. Second might be doable, second type of handlers are shell scripts. Need web UI to do change those, so any ideas are welcome how to disable service and host checks and also alarms.

1) I'm saying that for the forums, we're only supposed to give general advice, whereas when you open a case on Answer Hub, the support engineers will be able to do more that's specific to your environment and deployment (i.e. help with debugging any issues post-migration).

2) One other option, if you just want to disable notifications overall, would be to use this external command for the duration of time you want notifications disabled. You would just need to remember to run the inverse command later.