I have a node with 20% packet loss. Per ping graphs from NPM, the node is unreachable for up to 10 minutes at a time. I do not get any node warning or node down alerts for this device. Node warning interval is set to default 120sec.
If I open one of the availability graphs from the node view it reports 100%, but the default time interval is 30min. If I change the interval to 5min I do see gaps. Availability Reports also show 100% so I assume they use the 30min time period.
Two Questions:
1) Any reason this wouldn't be triggering a down node alert?
2) Is there a way to default the time period for Availability graphs and reports to 5min instead of 30min?
I currently have my Availability calculation set to method1. I have customers that view the Availability Reports and I would hate for a little packet loss here and there to skew my normally good numbers. I am concerned though that a site with a pretty consistent 10-20% loss is showing 100% available.
Thanks