i using cqlsstablewriter write sstables in offline/bulk mode. order not enforced during write operation. possible enforce compaction before use sstableloader load data cassandra cluster?
sstables immutable in nature, sstable not file having data metadata. meta data includes index.db etc. check datastax docs more details.
so should not manually token range in each sstable change during compaction , resultant sstable not having data evenly distributed. compaction leads larger sstable , node having sstable become hotspot.
it better/recommended not manually.
Comments
Post a Comment