How often does nagios poll the servers and where can i change how often nagios checks the servers and its services. Does it email immediately when some service is down?
offlate one of the services i was monitoring with nagios, went down and the email notification came about 10 minus later.
Can someone help me to the cause, as to why this could have happened.
Thanks:)
max_check_attempts 3
normal_check_interval 10
retry_check_interval 1
I suppose this should answer your quesition…
the host is checekd every 10 minutes, IF it fails it is checked again for another 2 times at 1 minute interval, if the third check fails a notification is sent.
It’s your configuration, not nagios.
Luca
Luca is right, you tell us how often nagios checks and retries until a notification is sent. We can’t see how you setup the .cfg files unless you tell us.
Thank you very much Sir Luca and Jake,
here is a part of my services.cfg file, which clearly indicates that my polling interval was 10 minutes, which is quuite high 
Sir, if i have to set it to 30 seconds interval, will the max_check_attempts be .60, does nagios recognize decimal point. Also is the email notification send immediately when there is a service interruption. Thanks
define service{
use generic-nix ; Name of service template to use
host_name TS01
service_description UPTIME
is_volatile 0
check_period 24x7
max_check_attempts 3
normal_check_interval 5
retry_check_interval 1
contact_groups Admins
notification_interval 120
notification_period none
notification_options c,r
check_command check_nt_uptime
}Sorry thats 5 minus and not 10
Another addition, my notification period is not none, it is set to 24X7
max_check_attempts 3
normal_check_interval 5
retry_check_interval 1
So the service is checked every 5 minutes, if in your nagios.cfg you have interval_length=60. You have a retry of 1 and a max check attempts of 3, so that makes 8 muntes. Sounds like it’s working just the way you have it defined, give or take 2 minutes.
But if what you really want, is to be notified approx in 5 minutes, then you need to define it that way.
max_check_attempts 1
normal_check_interval 5
retry_check_interval 1
per the docs “
max_check_attempts: This directive is used to define the number of times that Nagios will retry the service check command if it returns any state other than an OK state. Setting this value to 1 will cause Nagios to generate an alert without retrying the service check again.”
Use caution though, some service checks just might fail due to load at that given time on your network, timeout’s packet loss, etc. Not allowing check_ping a second chance one minute later, might cause you to get lots of notifications that have nothing to do with the service and everything to do with your network.
As of now i have put this,
max_check_attempts 3
normal_check_interval 1
retry_check_interval 1
Also the interval_length=30(seconds)
Does this mean every 30 seconds it will check. Does the notification come immediately when nagios sees any service down during that interval
Sir,
the scenario in my case is as follows. As soon as a particular service goes down, i need a warning signal as soon as possible, because that particular service is responsible for our uptime or downtime. Thats the reason, i am trying to keep a continuous check on that service so it can generate an alarm and email immediately to all responsible IT group.
thanks
Also the interval_length=30 means 30 seconds for each interval.
Since normal_check_interval 1 that means 1 interval which is 30 seconds. I don’t know why you would want to check something every 30 seconds.
retry_check_interval 1 which is 30 seconds.
max_check_attempts 3
So if a service fails it will retry the check 3 more times, each retry will be 30 seconds apart. You do the math, my head hurts.
I don’t know why you are doing this, but it’s excessive if you ask me.
60 second intervals makes for easy math.
retry_check_interval 5 means 5 minutes, nice and easy.
retry_check_interval 1 but since max_check_attempts 1 is how I would do it, but hey, whatever works for you.
Ah, now i get a more clear picture. Thanks
anyway as said by Jakkedup give it at least a second chance to cut off most false warnings.
Luca