Runtime Objects are missing

New satellites were added in the cluster … 2 HA masters and multiple zones with 2 satellite hosts in each zone. Zone configs got synced properly. But for some reason, downtimes and comments didn’t.

Any idea about the cause of this problem and a possible fix?

Would re-adding them would help to get them synced?

Thanks

Hello moniga!

Are there any unusual log messages during config sync between the involved nodes?

Best,
A/K

I didn’t see any unusual messages. I see only messages about the zone and global template sync messages.

So earlier these new satellites had higher version of Icinga on them and they are added. During that time, the ConfigObject tool messages were seen like creating downtimes on them Later on the icinga2 pacages were removed, compatible icinga packages wer installed and added in the cluster. After that they would not sync downtimes comments …

Do cache or icigna.state have to removed to have Icigna sync them?

Thanks

This can be tested easily: Create a new comment on a checkable in a satellite zone and watch whether it gets synced there properly.

1 Like

Thanks. I will test that out.

One of the satellites have /var/lib/icinga2/api/packages directory empty. I wonder if it will be create _api and stage directories?

Thanks

Ok adding a comment on service did the trick. I saw the directories got created and the comment .

Now the question is - why master wouldn’t sync the existing downtime it has to these satellites. Is there any way to make it do that or do I’ve to copy the downtimes conf files from master to them?

Thanks

I suppose master won’t do it considering it is runtime object. I will just add new downtime.

Thanks