How to use NPM to monitor network loop？

How would you identify the loop if you did not have NPM?

What network policies do you have in place that help prevent loops?

For example, do you have spanning-tree active? This stops the simplest of loops and many switches will report loops that are blocked with syslog and/or trap messages that can be used to generate alerts:

port A19 is Blocked by STP

Can users connect their own switches? Again syslog and STP is your friend

ESWD_STP_ROOT_PROTECT_IN_EFFECT: ge-0/0/6.0: root protect in effect for instance 0

Unfortunately some switches do not participate in spanning tree, and actually discard BPDU packets preventing loop detection. In that case we look for excessive ARP storm on an interface. syslog messages that look like this:

fpc5 Host Loopback:HOST LOOPBACK WEDGE DETECTED IN PFE 0

We use NPM to 'surface' these sorts of errors to the NOC staff so we can start troubleshooting faster,

If you use is not managed switch causing network loop? Whether through the upper switch to monitor whether network loop?

How would you identify the loop if you did not have NPM?

finding the never ending trace route, or seeing your ping's TTL (Time to live) drastically reduced.