Since you still have two SSTables after a major compaction, my best
guess is one of them is repaired and the other is unrepaired. If a
partition exists in both SSTable, and only one of them is a tombstone,
the tombstone must be kept to avoid resurrecting the data from the other
SSTable. To get
Hi all,
I have a table (with default gc_grace_seconds) where we have some row
tombstones created. Currently there are 2 sstables for this.
When we check the sstablemetadata, we found Estimated droppable tombstones:
0.689(for sstable1) and 1.094(for sstable2)
After executing nodetool compact on thi
