Cassandra is not that bad, maybe I didn’t operate it at large enough scale - I only had ~50 nodes - but u was doing 15,000 writes a second average, and 45,000 p

Cassandra is not that bad, maybe I didn’t operate it at large enough scale - I only had ~50 nodes - but u was doing 15,000 writes a second average, and 45,000 peak. I never really noticed the GC issues and only had to tune a single GC parameter once. The bigger issues I had was the need for all the extra disk space when a major compaction happens.