question

In my experience, the most common cause of high CPU utilisation on nodes is JVM garbage collection.

This is usually a result of:

• nodes being overloaded,
• disks not able to keep up with IO requests,
• or both.

You've provided very limited info but my best guess is that the commitlog disk is overloaded with writes because the node is overloaded.

Our general recommendation is to use SSDs on the nodes and where practical, to place the data and commitlog directories on separate disks (unless nodes have NVMe SSDs) to make sure that reads are not competing for the same IO as writes.

If your cluster has hit its maximum capacity, you need to review the size of your cluster and consider adding more nodes to increase throughput. Cheers!