Database Administrators Stack Exchange is a question and answer site for database professionals who wish to improve their database skills and learn from others in the community. It only takes a minute to sign up.
Sign up to join this community
I have 3 node cluster and upon checking the nodetool status; Load is just less than 100 GB on all three nodes. The replication factor is two and percentage own is 65-70% for all three.
However when I inspected the /data directory it is having index.db files for size more than 400 GB and the total size of the keyspace directory is more than 700GB.
Any idea on why a huge gap??
Let me know if any extra details are required :)
PS: nodetool listsnapshots command shows an empty list (No snapshots)
There are several reasons why there is a discrepancy between the data size reported by nodetool status and the output of Linux utilities such as du and df.
In your case where you think there is about a 600GB difference, the most likely cause is that there are files on disk which are not used by Cassandra usually because someone has placed it there. For example, someone copied files into one or more directories as an attempt to restore some data.
My suggestion is to review each table directory, comparing it's size on disk with the size as reported by nodetool tablestats to try to narrow it down. Cheers!
👉 Please support the Apache Cassandra community by hovering over the cassandra tag then click on the Watch tag button. 🙏 Thanks!
