Recently encountered a strange question:
On a new server, esxi 5.1, IP 136, added to the cluster, made a HA, using the EVC enhanced mode from the original host test to migrate a VM, test normal.
Because the server needs to be borrowed, it is removed from the cluster. First close inside of the VM, and then remove the host directly (no access to the maintenance mode, directly removed).
Over a period of time, the host returned, re install IP esxi5.1 136, but could not be added to the cluster!
Exclude as follows:
1 the client can connect 136 instructions to install the host.
2 network topology is as follows:
136--- direct --vice core -fiber- core ; core-- blade --san - storage ; 136--HBA- storage;
ESXi setup on blade, include vc5.1
3 from the core can be Ping 136, MAC address and the actual host
From the main core can not be Ping through 136, ARP display MAC address error, through the optical fiber dynamic Dynamic from the blade to learn!!!
But there is no MAC address on the blade, and also the ARP cache, the problem remains.
4 VC view, you can ping through 136, but the MAC address is wrong, and same as the main core swtich
Now the problem is: 136 and there is no direct knife box, why MAC address is wrong? Which sends the MAC message?
