I’m experiencing what I think may be due to something not quite configured as efficiently as it could be.
I’ve been building up a number of checks for a series of around 250 servers, all of which work fine. Now having built all these checks, it was time to make some of these service checks dependent upon others succeeding (which they were in fact). I do not get any errors in any of my config files, but I have noticed something that is a little distressing.
I add one service dependency, it takes my pre-flight check and the nagios restart 2 minutes to complete.
I add two service dependencies, it takes my pre-flight check and the nagios restart 5 minutes to complete.
I add seven service dependencies, it takes my pre-flight check and the nagios restart 1 hour and 10 minutes to complete.
NOTE: I also placed all seven dependencies in their own servicegroup, but call out each service in the service_dependency configuration entry. I don’t know if this is of any significance or not.
Every time I run my pre-flight it runs through all the checks quickly until it his the following line…
“Checking for circular host and service dependencies…”
I also noticed that the “Checked xxxx service dependencies.” line had over 2000 checks with one service dependency, over 6000 with two, and over 90,000 with seven.
Does it really take over an hour to check seven service dependencies that are dependent on only one other service across 160 of 250 servers? Or am I doing something bad that isn’t fatal, but is time consuming?
Thanks in advance.