Host disconnected

[pvhlarson]

We are running 2 hosts in a cluster, over the weekend I accidently gave the ip of one of theVsphere Center to another server on the domain causing an ip conflict preventing access to both servers. We got it sorted out eventually, but in the process had to change the folder name of the Vsphere Center vm.

All machines are functional, but now in the cluster one of the hosts is grayed out and disconnected, if I try and connect I get an error: Cannot contact the specified host. The host may not be available on the network, a network configuration problem may exist or the management services on this host may not be responding.

How do I fix this and can it be fixed without impacting the vms on the host?

I can connect to the host through ssh and the client.

Thanks in advance.  Peter

[dales123]

How are you trying to connect the host are you using the IP or DNS name for the connection, if using DNS check that pinging the hostname replys with the expected results if not try a different method of connection so if your using IP at the moment try changing to DNS and vice versa.

Have you restarted the management agents on the host just to make sure they have not got themselves in a spin.

Also I take it nothing has happened to the storage at all at the same time as this also could be a symptom of All Paths Down but you'll want to check the /var/log for some further hints on what the problem actually is if its not something straightforward .

[a_p_]

We got it sorted out eventually, but in the process had to change the folder name of the Vsphere Center vm.

I assume you only changed the folder name, but the IP address remained the same!?

Did you double check whether the DNS name resolution for the vCenter Server works correctly? Are you able to ping the vCenter Server from the host by it's short name (host name) as well as with the FQDN?

Unless you have a special configuration on the host, try to restart the Management Network from the DCUI (yellow-gray console).

André

[pvhlarson]

we are using ip for the connection, it responds to ping of the ip fine, we have restarted the management service and it did not improve things. Thanks, Peter

[dales123]

Ok,  Are you sure that the hosts management IP address has not also been duplicated as well elsewhere if you ping the management IP from a desktop and then do an arp -a does the mac address registered to the IP match that of the ESXi servers management port, if not then the IP has been duplicated. Try giving it a different management IP (thats definately not being used) and see if you can ping that. if its not working at that point your only real hope is to scour the log files and see if there is anything interesting in there, if not you'll have to take down the vm's and reboot the host to try and rectify the connectivity.

[pvhlarson]

After a couple of hours digging with VMWare support we saw the MAC address on the NIC was showing up differently in arp cache depending where you ran it, turned out someone had given a DRAC the same IP address on the subnet and it for some reason it picked that weekend to wreak havoc, it had been running fine for the 6 months it was set up that way, so we're not sure what triggered it to go off when it did.

Thanks for all the input.  Peter