Hello everyone,
I'm seeing a weird thing on a C4948E running the latest IOS (15.2(4)E10) whose interfaces are cacti-monitored - there is one SVI ("counter" is enabled in its config) whose graphs show constant ~15M in and ~15M out of unicast pps; since that's a "back-up" interface (runs OSPF with high cost configured, no other traffic) I pulled the counters manually and noticed this:
$ snmpget -v2c -c public X.X.X.X ifInUcastPkts.137; sleep 10; snmpget -v2c -c public X.X.X.X ifInUcastPkts.137
.1.3.6.1.2.1.2.2.1.11.137 = Counter32: 913129909
.1.3.6.1.2.1.2.2.1.11.137 = Counter32: 913129903
Yes, the counter gets decremented over time (and therefore cacti assumes it overflows constantly). If I generate some traffic (for example ping the OSPF neighbour) the counter increases...
Has anyone encountered this sort of behaviour?
Thank you.