Here are my notes on the upgrade procedure:

Step 11a:  Before restarting cloud-management, make sure to chown cloud:cloud /var/lib/cloud as per "Known Issues 3.2"

Step 11b:  Step instructs to "wait for the database upgrade to be completed" however doesn't give any instructions on how to confirm this.  To confirm the database upgrade is completed, you should look for the following lines in your management-server.log file:

INFO  [cloud.upgrade.DatabaseUpgradeChecker] (main:null) DB version = XXXX Code Version = XXXXXXX
 

INFO  [cloud.upgrade.DatabaseUpgradeChecker] (main:null) DB version and code version matches so no upgrade needed.

Of course, the expectation is "DB version" and "Code version" should match the version you've upgraded to.

Step 16: Use 'nohup cloud-sysvmadm -d 192.168.1.5 -u cloud -p password -a > sysvm.log &'.  (replace 'cloud' and 'password' with appropriate username and password)

In my case, step16 failed to restart one of the routers as well as one of the sysvms.  Keep an eye on the log.  If it reports failing to restart one of the VMs, stop and start manually.