Jon Haddad

Most database performance tuning starts with assumptions and ends with hope. I took a different approach: build disposable lab environments, run systematic experiments, and capture what actually works.

In our talk, we introduce two new tools designed to streamline the deployment, testing, and exploration of Cassandra clusters regardless of version. "easy-cass-lab" simplifies the setup of Cassandra clusters, incorporating built-in profiling utilities to generate flamegraphs and analyze system performance with eBPF, while "easy-cass-stress" facilitates the creation of workloads to test against the

Despite significant enhancements in CPU cores, system memory, garbage collection, and disk technology, the Cassandra community maintains conservative node density standards. This talk examines the disconnect between hardware capabilities and storage practices in Cassandra environments.

Apache Cassandra is renowned for its remarkable scalability in the face of a rapidly expanding data footprint. However, to fully harness its potential, it is essential to equip yourself with a repertoire of invaluable techniques. Drawing from two-decades of experience in database optimization, our s

There's a direct correlation between database performance and how much it costs to run. If you're using the default Cassandra and OS configuration, you're not getting the most out of your cluster.