I have a cluster which I am considering enabling incremental repair on. If anything goes wrong I'd like to disable incremental repair on every node. How do I do that?
Turn node off and use sstablerepairedset to remove the repair time for each sstable so that they will all be candidates for future compactions.
find '/path/cassandra/data/keyspace/table/' -iname "*Data.db*" > sstables.txt
sudo -u cassandra sstablerepairedset --is-unrepaired -f sstables.txt
Then just go back to using repair with no -inc or in later versions use the -full flag