We have Cassandra version 0.7.5 with 8 node cluster.
The Cassandra is located on server with 24GB and 8 cores CPU.
Each day the Cassandra write 10 million rows every 6 days we delete the last day.
We want to run a compact maintenance, today every day we run Repair on each node and after that Clean.
While we running Repair and Clean we've noticed that the database is grow too much.
After the Repair and Clean finished we can see that the database growing almost double.
What are the steps to run the preparation for the maintenance work?
What are the commands that need to be run to compact the data and save space?