Clear connections doesn't solve the actual issue. You should try to understand why there is a spike in the traffic . Please do perform a debug on the edge interfaces and you will certainly see what kind of traffic is flowing via each interfaces which will root cause the issue up-to an extent. NSX Command Line Quick Reference
If the traffic is coming from different DC farm or MPLS you will need help from network team as well. If you have VRNI configured in this setup , it will be very easy for you to know the flows.
I think if you deploy vRealize Network insight and integrate it with NSX, you can have a deep insight into all the traffic you have in your DC specifically the North-South traffic passing through the NSX edges with all sources and destinations. This way you can see if these are normal or not.
Ok, I will install vRealize Network insight and try to figure this out.