I have an ideal for backups in my mind with Cassandra to dump each columnfamily to a directory and use an offline process to compact them all into one sstable (or max sstable size set). I have an ideal for restoration which involves a streaming read an sstable set and output based on whether the data fits within a token range. The result of this is that I can store a single copy of data that is effectively already repaired and can read from the specific range that covers a node that I wish to restore. My first look at this was somewhat frustrated by sstable code in the current versions have a strong reliance on the system keyspace.
Does anybody have any thoughts in regards to other things that might exist and fulfill this (particularly offline collective compaction), have a desire for such tools, or have any useful information for me before I attempt to build such beasts? -Jeff
