As I wrapped up the migration from vSphere 4.0 to 4.1, I had one host that for the life of me wouldn’t upgrade. It kept giving me the following error:
Check the network connectivity and firewall setup, and check esxupdate logs for details.
I can see the ESX host just fine. Nothing out of the ordinary there. The firewall ports look exactly the same as the other 6 hosts in this particular cluster. Why the hell won’t it download the updates.
Well, time to roll up the sleeves and log into the esx host with ssh. It didn’t take long for me to realize the issue.
PING cdalnvci01<snip> (10.6.1.152) 56(84) bytes of data.
64 bytes from cdalnvci01<snip> (10.6.1.152): icmp_seq=1 ttl=127 time=0.255 ms
64 bytes from cdalnvci01<snip> (10.6.1.152): icmp_seq=2 ttl=127 time=0.242 ms
Crap! Old IP address. Checked the /etc/hosts file and sure enough, there was an old entry sitting there mocking me. As soon as I commented out that line, the update manager was able to communicate with the host and all was well again.
I’m sure that there are others out there that at one time had a goofy setup so if this bit me, I’m sure it might bite you. Especially those of us that walk into a clients site to do some consulting and lord knows what all they have hacked in to get things to “work”.