Cassandra control SSTable size -

is there way control max size of sstable, example 100 mb when there more 100mb of data cf, cassandra creates next sstable?

unfortunately answer not simple, sizes of sstables influenced compaction strategy , there no direct way control max sstable size.

sstables created when memtables flushed disk sstables. size of these tables depends on memtable settings , size of heap (memtable_total_space_in_mb being large influencer). typically these sstables pretty small. sstables merged part of process called compaction.

if use size-tiered compaction strategy have opportunity have large sstables. stcs combine sstables in minor compaction when there @ least min_threshold (default 4) sstables of same size combining them 1 file, expiring data , merging keys. has possibility create large sstables after while.

using leveled compaction strategy there sstable_size_in_mb option controls target size sstables. in general sstables less or equal size unless have partition key lot of data ('wide rows').

i haven't experimented date-tiered compaction strategy yet, works similar stcs in merges files of same size, keeps data in time order , has configuration stop compacting old data (max_sstable_age_days) interesting.

the key find compaction strategy works best data , tune properties around works best data model / environment.

you can read more configuration settings compaction here , read this guide understand whether stcs or lcs appropriate you.