This is on single node environment, and there is a lot of data in commit logs needs to be flushed to sstables. So remove commit logs will cause data lost.
Fortunately, this issue happens on a new key space that have CFs that use composite type, and we can sacrifice loosing data on this new key space. So we modified cassandra code so that cassandra will continue load when it encounters this issue. We was able to get cassandra up and running with the modified code, and data in other key spaces remain intact. The schema change was that we created a new key space with composite type CFs, but later we had to change some definition/CF names, so we dropped the key space and recreated with new definition. Huy -- View this message in context: http://cassandra-user-incubator-apache-org.3065146.n2.nabble.com/Cannot-start-cassandra-node-anymore-tp7150978p7227283.html Sent from the cassandra-u...@incubator.apache.org mailing list archive at Nabble.com.
