SQL cluster update - Failover to un-patched node

So when updating a two node cluster, our procedure is this: 1. Backup all databases. 2. Failover all SQL cluster roles to Node A. 3. Install the patches on Node B (passive). 4. Restart Node B. 5. Failover all SQL cluster roles to Node B. 6. Install the patches on Node A (passive). 7. Restart Node A My understanding is that in step 5, the updated node, will perform some updates before taking on incoming connections, as can be seen in the log. What happens if another failover occurs before Node A is updated? If, for example, I have updated to a higher CU on Node B. Will the failover fail or just fail over and start itself on a lower CU level?