Hi All,
Hoping someone has some insights into a strange issue I’m having (hopefully a simple issue that I’m not seeing…).
I have 2 pfSense appliances with LAN addresses on a /24 network:
Appliance-1 : 10.250.1.102
Appliance-2 : 10.250.1.103
There is no HA pairing (yet).
Onto these I have created 2 CARP interfaces:
10.250.1.100
10.250.1.101
What I am looking for is for Appliance-1 to claim Master for the .100 address and Appliance-2 to claim Master for the .101 address.
The CARP addresses have been created identically on both appliances with the exception of the Skew (screenshots from Appliance-1, but the config is the same on appliance 2, except the reverse Skew on each VIP):
Appliance-1 VIPs:
So far so good - Both VIPs are created and respond correctly. Appliance-1 is Master for .100 and Appliance-2 is Master for .101
If I enter persistent CARP Maintenance Mode on Appliance-1, Appliance-2 takes over .100 and responds correctly. The same applies if I enter CARP maintenance on Appliance-2 : Appliance-1 takes over .101 and all is good.
The issue is if I shut down Appliance-1, Appliance-2 shows Master for both VIPs (as it should), but traffic to the .100 VIP is patchy at best. A simple ping shows is responding to only about 1 in 4 packets. This behavior is the same if I shutdown Appliance-2. Appliance 1 claims Master over the .101 VIP (now being Master for both VIPs), but only responds to occasional pings.
For completeness, these are virtual appliances running on ESXi. The port group they are attached to have security settings enabled to allow promiscuous mode, MAC address changes, etc, and works for other CARP servers on the same subnet.
Any insight would be greatly appreciated!
Thanks,
Sam