First of all, apologies if this has been asked before, I cannot find an answer on Google or StackExchange.
Is it possible to put Apache Cassandra into a split-brain scenario where a network partitioned data service continues to run storing and accessing data without access to the full cluster?
If not, what are the techniques used (links to documentation OK) to avoid a split-brain?
If so, (there seems to be tweets about this happening) what are the recovery options in this case?