This project involved designing and deploying a high-availability distributed data platform using ScyllaDB to support scalable, low-latency transactional workloads.
The architecture consisted of a 6-node ScyllaDB cluster deployed across two datacenters to ensure resilience and disaster recovery. The cluster was configured using NetworkTopologyStrategy with a replication factor of 3 per datacenter, with three nodes located in the primary datacenter and three nodes located in the disaster recovery datacenter.
This setup ensured that data was replicated across multiple nodes within each datacenter while also maintaining cross-datacenter redundancy. The configuration allowed the platform to maintain service availability even in the event of node or datacenter failure.
To balance durability and performance, the system used LOCAL_QUORUM consistency for read and write operations, enabling low-latency access from the primary datacenter while maintaining strong data consistency. Replication between datacenters ensured that the disaster recovery site always maintained up-to-date copies of the data.
The platform was designed to support high-throughput application workloads, including payment processing and backend service transactions, with optimized shard-aware performance and distributed data partitioning.
Monitoring and operational visibility were implemented using cluster metrics and observability tools, enabling proactive detection of performance issues and ensuring stable database operations.
The final architecture provided:
Ensure that we have these nodes on cloud and utilise the cloud autoscale
