[Nagiosplug-devel] time problem with passive service checks

[administrator at net-and-works.de]

Hi,

first, there is an inconsistant description of freshness checks in the
documentation.
the freshness_threshold must be given in seconds, but the freshness.html
says: if set to 5 and interval_length is set to 60 then it will be 5
minutes. this isn't true. see paragraph "How The Freshness Threshold Works"
the same wrong description is given on the page distributed.html paragraph
"Freshness Checking".

second:
I'm providing some passive service checks to nagios without using nsca. the
check time is provided by the remote host. the check is executed every 5
minutes and freshness_threshold is set to 10 minutes. the problem is, that
the remote hosts time isn't synchronized with the nagios hosts time. if the
remote hosts time is more then 10 minutes after the nagios time, then the
service states will flap all the time, because each incoming passive
service check result is detected to be stale. this could be solved by
synchronizing the times with NTP, but which way should be used if NTP isn't
possible. How does NSCA handle this?
As a workaround i do not use the check time from the remote host, but
replace it with the local time when passing the result into the external
command file. any suggestions about a better solution?

martin