Microsoft Multicast NLB causes network disruption to VPN clients
Hi,
I've just deployed a Microsoft Multicast NLB cluster of two Windows Server 2022 nodes onto my ESXi 7.0 U2 host which has a two NIC etherchannel connection to my Cisco 9200L switch.
I have VPN clients served by Microsoft RRAS, with a pool of 10.0.0.50-10.0.0.80.
This all sits in my subnet of 10.0.0.0/23.
I find that if any device tries to access the NLB cluster via any service other than ping, clients on the local network are fine, but it causes any connected VPN clients to have their network connectivity temporarily disrupted.
I can't think of any reason these VPN clients specifically are affected and would assume if the Multicast traffic is somehow causing high CPU on the switch or port flooding, even local clients would be affected.
Any thoughts what's occurring here or how to go about diagnosing why these clients are dropping?
Thanks
James