How do you handle network partitions and maintain data consistency across a distributed storage cluster?
