Investigate how unreplicated data is transferred in a planned failover
Problem to solve
We currently state in our Planned failover documentation that in order to copy unreplicated data, a user must use some form of
rsync or backup strategy.
At this point it is not clear what this entails for each data type that Geo doesn't support. We should investigate the process and ensure we understand current limitations.
In this form the documentation is not very helpful
Investigate how all unreplicated data types can be transferred at this point and if there are any known limitations. This should improve our existing documentation and may help us identify bottlenecks in the planned failover procedure.