When testing HA functionallity, disconnecting the backup/redundant Service Console may cause the ESX host to be uncontactable in VirtualCenter although the primary Service Console IP address is still contactable from hosts on the same subnet as the primary Service Console.
This would appear to be due to a routing issue in the service console. e.g.
# esxcfg-vswif -l
Name Port Group IP Address Netmask Broadcast Enabled DHCP
vswif0 Service Console 172.16.1.10 255.255.255.0 172.16.1.255 true false
vswif1 Backup Service Console 172.16.2.10 255.255.255.0 172.16.2.255 true false
# route
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
172.16.1.0 * 255.255.255.0 U 0 0 0 vswif0
172.16.2.0 * 255.255.255.0 U 0 0 0 vswif1
default 172.254.1.1 0.0.0.0 UG 0 0 0 vswif0
VirtualCenter is on the 172.16.2.x network.
To resolve the issue either add a route or use an IP address on a different subnet from VirtualCenter and different from the Service Console.
e.g. Add a route for the VirtualCenter network to the primary Service Console:
route add -net 172.16.2.0 netmask 255.255.255.0 vswif0
Changes can be made permanent by editing /etc/sysconfig/network-scripts/route-vswif0 with the following entries:
ADDRESS0=172.16.2.20
NETMASK0=255.255.255.255
GATEWAY0=172.16.2.1
Test this updates the routing table with either of the following commands:
/etc/sysconfig/network-scripts/ifup-routes vswif0
service network restart
A better solution is to use a separate subnet from the primary Service Console and from VirtualCenter for the backup Service Console e.g. 172.16.3.x
Do not use the same subnet for the primary Service Console and the backup Service Console as this will not function as expected:
VMware ESX High Availability not working as expected
http://theether.net/kb/100041