Before upgrading your environment, one thing that can help insure that you have clean data prior to any upgrade is to verify that there are no issues present in the system due to anomalies in processing (server communication loss, network issues, dropped connections, etc.) that may have resulted in some oddities in data.
Another recommended procedure that we are happy to see being utilized heavily, is the practice of using DEV and UAT environments to learn to crawl > walk > run with the practice of upgrading your instances. This does allow you to pre-determine and set appropriate expectations on timing of the effort, and work involved. If however, you should find yourself concerned about particular tasks or upgrade performance, please contact support, as we can and will assist if there are issues encountered and will be happy to help you.
