Can't reach some hosts after upgrade to 1.6

[tmashos]

I’ve got a really weird issue after upgrading to 1.6.0p9 from 1.5.0p26 (I think). Some hosts are unreachable and specify an error that it cannot be resolved in DNS. This is true, as I specify the IP address for all my hosts.

The issue is similar to this, however it’s not all hosts and I’ve already fixed that all_hosts statement.

[Check_mk (english)] no hosts after upgrade to 1.6p2

Hi, I have tried to upgrade my 1.5p22 to 1.6p2 but now I don’t have any hosts. I can see them in WATO, but when I try to edit the services, I get the error “Failed to lookup IPv4 address of via DNS: [Errno -2] Name or service not known”. This seems clear to me, because the hosts are not resolvable via DNA. That’s why I have entered the IPs for each host, but Check_mk seems to ignore the IP and tries to resolve the name via DNS anyways. Has anyone got the same problem and possibly a solution? …

Some background, I use some automation tools (puppet) to generate the hosts.mk file in the new format specified here Changed format of host tags in hosts.mk configuration files (I updated this from the old format).

The hosts in question are not found when searching in the quick search, however I can drill down in WATO to find them. Some of the hosts are part of a cluster and I get an error now that they aren’t monitored on this site.

Now here’s where things get kinda weird.

Not all hosts in the same folder (same hosts.mk file) have this issue.

1. If I attempt to query the host from the command line via cmk --debug -vvn HOSTNAME it queries the host correctly and I can see the state of all the services.

2. If I edit any host in that same folder, but just hit save without making any changes, all the hosts are now reachable (Essentially it’s only changing the format of the file)

I’m not sure what else to look at for this or what else to dig into.