I'm having the following problem: I'm running two Cumulus Linux (Edgecore 5812-54X-O-AC-F) switches in a VRR setup, a couple of access switches (also CL) with dozens of clients (Windows, Linux, voip). In a few vlans/subnets I'm seeing the following issue.
A client sends out an arp request for its default gateway, gets a response and is able to ping to anything. Then after a few seconds no ping replies are received anymore and the VRR switches show a STALE entry on ip neigh show. One of the VRR switches will still respond to arp requests. Whenever the client reconnects the same behavior is shown. Does anyone have an idea what the underlying issue could be? The behavior is not apparent on all vlans/subnets and the clients affected seem to be at random.
I've seen the following behavior on the VRR switches regarding their ip neigh entries: DELAY, PROBE and then they go into STALE on both switches. Other, working, clients show a REACHABLE state. Any pointers are appreciated!
No comments:
Post a Comment