I'm using NPM 9.5.1 and have configured the following alert for a switch stack on SNMPv2:
Trigger when ALL the following apply:
- Node Name is equal to xxxx
- Interface Status is equal to Up
- Interface speed is less than or equal to 100000000 (100mbit)
Check every minute, trigger condition must exist for one minute.
I tested it today by plugging a laptop into the switch stack and hardcoding different 10mbit and 100mbit configurations from both laptop and switch.
Polling interval for the switch is 120 seconds, rediscover is 30 minutes, and statistic polling is 9 minutes for the interface/5 for the node.
NPM did not recognize that the port had negotiated to a lower speed, even after numerous forced polls and rediscoveries of the node. I noticed the interface speed finally update around the time I manually polled/rediscovered the interface itself, which was over half an hour past when I originally lowered the interface speed.
Why did it take so long? Does device polling check interface speeds? Or does rediscover do that? I need the alert to fire immediately, or within two minutes at least - it's a critical switch stack. Not sure about the validity of rediscovering an interface that often.. I thought that's what polling was for.
My apologies if this is in official documentation, I hadn't found any in my searches so far. Thanks in advance for any ideas on what I'm doing wrong here.
