
Originally Posted by
esilverst
We would like to test our 3.2 binaries on a 3.2 and 4.0 cluster where the master will reside on the same machine. Any potential issues with this?
The reason for this question, is we have a client who wants to upgrade to 4.0, but are heavily using their 3.2 cluster for some long running tests. They don't want to get any downtime from the 4.0 upgrade. So we'd like to get both clusters running concurrently, and switch them over once the 4.0 cluster is operational. Most importantly, they don't want their long running tests to fail because of the upgrade.