This is a project revolved around system design for a booking system. Both PostgreSQL and Apache Cassandra are benchmarked against one another (with the ultimate optimizations being for a Cassandra cluster).
For the Apache Cassandra design, I chose to implement a single rack, 9 node cluster in my local datacenter (us-west-1) onto AWS.
Add row caching to significantly increase read heavy operations Add a data tiered compaction strategy Add Compression Horizontally scaling with an Nginx load balancer Vertically scaling by adding new nodes to cluster gradually
WRITE: increased from an intial throughput of < 60 rps to throughput of > 1200 rps with an error rate < 1.5% READ: increased from an initial throughput of < 75 rps to throughput of > 1500 rps error rate < 1%