polling for slow connections

[michaewlewis]

I have a bunch of remote telemetry sites connected via 450MHz radios (point to multipoint) that get overwhelmed with much more than 5kbps throughput (800bps is typical). Is there a way to make Nagios aware of slow network connections and spread out requests to avoid overloading the connection?

I had an issue when I started grabbing snmp data where multiple retries started giving me a denial of service, which led to more and more retries. To fix that, I took retries to 60 minutes with a max attemps of 1. But from looking at the scheduled checks for these sites in Nagios, the checks are pretty random, but there are still some checks that are happening within a few seconds of each other. On a normal network, that wouldn't be an option, obviously, but when I'm limited to a few kbps, and responses average 1500ms, it turns into an issue.

So, basically, I'm wanting to know if there is a way to control how close together requests are to hosts or groups of hosts and their services.

[cdienger]

Check out the service_inter_check_delay_method and host_inter_check_delay_method options available for the nagios.cfg at https://assets.nagios.com/downloads/nag ... gmain.html. Both of these are set to 's' by default but you can specify a time(in seconds) to space events. You may need to also disable the auto_reschedule_checks option - I've not tested these options before so I'd start by setting the delays first and tweak the reschedule option if necessary. The nagios.cfg can be edited through the web UI under Configure > Core Config Manager > CCM Admin > Core Configs. Makes sure to save changes and then apply configuration if you make changes to the file.