One thing that might cause this is a difference in the timescales of Netflow reporting vs. interface polling.
We ran into this once on interfaces that had a lot of long-lived flows (e.g. tunnels) and we had left the Netflow active flow cache timeout at its default value, which on our devices was 30 minutes. So a signficant portion of the Netflow traffic was being reported in bursts at 30 minute intervals.
SolarWinds solutions are rooted in our deep connection to our user base in the THWACK® online community. More than 150,000 members are here to solve problems, share technology and best practices, and directly contribute to our product development process. Learn more today by joining now.