I've been having problems with nagios checks reporting "No Information Returned From Host Check" when reloading the config via "/etc/init.d/nagios reload" I have fixed this problem by changing the init.d script to not use the killall command when reloading the config, but to use the kill command instead. I now have the following in my init.d script and it works fine. reload() { checkconfig || return 1 if [ -f /var/nagios/nagios.lock ]; then ebegin "Reloading configuration" kill -HUP `cat /var/nagios/nagios.lock` &>/dev/null eend $? fi } I believe that the old way of doing things i.e. using "killall -HUP nagios" was also killing the nagios children doing checks and they were reporting nothing back to the main daemon. I hope this can be included into the next release :-)
A nicer way of doing this is as follows: reload() { checkconfig || return 1 ebegin "Reloading configuration" [ -f /var/nagios/nagios.lock ] && kill -HUP `cat /var/nagios/nagios.lock` &>/dev/null eend $? } This will also display a status output of [!!] when you've tried to reload the config when Nagios isn't running!
ok, this one is confusing me now as its started to fail to reload the config :-( in /var/nagios/nagios.log we have the following last entry: [1162910170] Caught SIGHUP, restarting... It seems that a kill -HUP kills the daemon rather than restarting it. I've tried executing the daemon form the command line and still the same thing happens: # /usr/nagios/bin/nagios -d /etc/nagios/nagios.cfg # kill -HUP `cat /var/nagios/nagios.lock` when I then look in the process list, there are no nagios items listed :-(
which exact version of nagios are using?
the masked 2.5 version.
Has become a bit stale, is this still an issue for you? Please reopen if so
