While upgrading from 7.0.x to 8.0.3, an error occurs during stage 2, Step 1, the following alert is displayed:
A problem has occurred. The source vCenter Server might have been Powered Off during this process.
1. Copy data from source vCenter Server
(The progress bar stops at 50%)
A problem occurred while - Applying network configuration...
The Messages button displays the following errors:
Failed to set the network
Error setting network. Details <Date><Time> [ERROR] Cannot run /sbin/ifup eth0 command.
Unknown error. Return code: 256 output: Make sure the interface is down or not assigned anyIP eth0 is DOWN or not assigned an IP. Bringing eth0 up... Can not find the manual filename, let us search for the auto filename Performing duplicate address check for IPv4 address ###.###.###.###:[1;31mERROR: IP already exists in network:[0;39m Unable to set network parameters
7.0.x
This was cause by entering the current vCenter FQDN into the FQDN section on step 9 (Configure Network Settings) of Stage 1 of the upgrade. This causes the new vCenter to query itself for the IP configuration instead of going through DNS and querying the old vCenter.
This can be proven by running a curl command from the new vCenter to the FQDN of the old vCenter.
# curl -v telnet://<vCenterFQDN>:443
* Trying 127.0.0.1:443...
* Connect to 127.0.0.1 port 443 failed: Connection refused
* Failed to connect to <vCenterFQDN> port 443 after 0 ms: Couldn't connect to server
* Closing connection 0
On the new vCenter, the hostname command returns the old vCenter FQDN.
The curl command succeeds when telnet-ing to the IP address of the old vCenter:
# curl -v telnet://<vCenter_IP>:443
* Trying ###.###.###.###:443...
* Connected to ###.###.###.### (###.###.###.###) port 443 (#0)
In order to avoid this issue, leave the FQDN section blank. The name will be moved to the new vCenter during stage 2 of the upgrade.