ping threshold settings?

[tomtom]

Good morning and thanks for all of your help with this. There was a bunch more notifications last night and the one noisy host is about to throw more errors:

rta=5000.000000ms;3000.000000;5000.000000;0.000000 pl=100%;80;100;0

I can access the system fine, but it does appear we are having network problems as well with random systems. Thanks!

[jdalrymple]

Were we able to help you adjust the thresholds properly and if so does that solve your Nagios specific issues, or should we leave the thread open for further discussion?

Happy to help.

[Box293]

From what I can see here, you're just exceeding the thresholds.

You could try something like:

Code: Select all

check_command check-host-alive!-w 10000.0,80% -c 15000.0,100% -p 5

[tomtom]

Hello again,

Yes, I've been slowly increasing the threshold. Still got warnings last night on 3 of the systems with the exact settings you specified so I just bumped it up to

check_command                   check-host-alive!-w 20000.0,80% -c 25000.0,100% -p5

So would enabling flap detection alleviate this at all. It's just strange that the ESXi host is complaining and only 2 of it's 20+ vm's are as well.

[jdalrymple]

20 and 25 second ping thresholds are pretty ridiculous. I'd almost consider that system down. I would speak to your network folks and have them investigate the slow ping replies between these 2 hosts. I can get from my site in MN to a site I work north of the Arctic Circle (which of course involves space travel) in under a half-second.

Flapping will only reduce alerts, not eliminationg them. Flapping would turn UP/DOWN/UP/DOWN/UP/DOWN/UP/DOWN/UP into UP/DOWN/FLAPPINGSTART/FLAPPINGSTOP.

Make sense?