I have about 10 IP SLA operations (voip udp jitter) running on a Cisco ISR4k router. Up until Monday it experienced only brief packet loss, and always below 5%. A change was made to our voip infrastructure on Saturday, which pushed most of our inbound/outbound calls through this router.
Starting Monday we began seeing high packet loss on our IP SLA operations sourced from this router, which probe to other routers. The packet loss is being reported at 20-80% throughout the day, stopping at night, and resuming the next morning when people show up in the office.
It definitely seems as if the change made has impacted our monitoring, but I'm not seeing higher cpu or memory utilization in NPM. I'd like to be able to correlate the packet loss we're seeing to a CLI output to either prove or disprove the operations are being affected. When I run "show ip sla statistics <operation number>", I only see tail drops and packet skipped counters, though the packet skipped does not increment. Tail drop seems to fluctuate between 0 and as much as 600-700. The CLI output shows no packet loss whatsoever no matter how many times I check.
Am I looking in the wrong place? How does Solarwinds determine packet loss here if the Cisco CLI doesn't show any?