What is off heap memory? How and why is it used for storing bloom filter, row cache, key cache and partition summary?
I was reading the read path on the documentation. During the read path on cassandra , read starts by moving through several stages like row cache, bloom filter. In most of these stages, the subsets of data are stored in memory "Off heap memory". I just need to know what is it? How is it storing the subsets and why not to store it in disk where the stables are being stored? What is the benefit of storing in off heap memory?