nagios 1.2 is out, i would love to see an ebuild for it. thanks!
Now that netmon herd let's put the bug there.. (ok, so it's still me fixing it but anyway..)
Ebuild for nagios-core 1.2 comitted to CVS and should hit your local rsync mirror in 30 minutes or so. Please test it and send us your feedback.
before testing, i'd like to know: you have updated nagios-core to 1.2 but not nagios to 1.2. is it normal? (another thing: nagios-1.1.ebuild has no digest).
How is nagios-core 1.2 working for you? BTW: just commited net-analyzer/nagios-1.2 (the meta package) into ~x86 as well..
I'll give this a test and report back since this bug hasn't recieved any comments. I have Nagios monitoring numerous machines, so I can provide a good stress test.
Two days with no problems so far. I did accidentally compile with the "postgres" useflag at first though, and noticed that it did not create the nagios binary. For normal usage however, I haven't run into any problems yet. I'll report back once more, next week.
I hadn't run into any problems at all until I upgraded to 2.6. Now, for some reason, the web interface cannot find the daemon process in order to send it commands. The daemon is running and sending out status e-mails still, the status pages are updating with changes, but you cannot issue commands via web interface. The PID is properly stored in /var/nagios/nagios.lock, which is owned by root:root, but the o+r bit is properly set. Changing its ownership to nagios:nagios doesn't remedy it. There is a slight possibility that this may not be a 2.6 problem and that I simply failed to notice all of this time, but it's marginal.
The commands from the CGIs to the nagios daemon is send through a pipe, called nagios.cmd. Can you check that you have such a file, and its permissions? If you have it, try to remove it, nagios is supposed to recreate it at restart.
