Impact of Cisco DNAC Integration on ISE Resource Utilization

[MSN]

After the integration of ISE with DNAC (SD-Access Deployment), I noticed memory spikes and high resource utilization in Cisco ISE. I would like to understand the impact of this integration on resource consumption. Specifically, clarification on the following:

1. What specific types of communication occur between Cisco DNA Center and ISE post-integration?
2. How does this traffic contribute to system resource utilization on ISE?
3. Are there known cases where pxGrid-related traffic or other API interactions from DNA Center cause significant resource spikes in ISE?

[Arne Bier]

I don't know if Cisco has published any document that tells you exactly what goes on between these two systems, or how much data is exchanged.  My advice would be to run a tcpdump on both ISE nodes that have pxGrid enabled - and set a filter to only capture the traffic to your DNAC (probably the Enterprise VIP, or whatever your setup looks like).  Run that for 60 minutes and then run it through Wireshark to see IO stats and other cool charts.  It will tell you everything you need to know. The ISE tcpdump filter to capture only the traffic to and from 10.1.2.3 looks like this:

ip host 10.1.2.3

.