I would like to have a single node cassandra cluster initially but to maintain backups for single node how about occasionally & temporarily adding a second node (one that would contain the backup, this could be my dev machine as well, far far from first node in some remote datacenter) to cluster as a replica so that data would be synchronized on both as if it were a replica.
Would it be possible to do this ? May be I could do this backup once 2-3 days.