Distributed Event-Driven Backend Experiments
A high-throughput telemetry ingestion and query pipeline prototype built on Cassandra and Java concurrency primitives.
Developer Tools, personal project
Problem and motivation
High-frequency device telemetry generated write bottlenecks in traditional relational databases.
Solution and architecture
A distributed event ingestion prototype uses Apache Cassandra for horizontal write scalability and custom Java worker pools for batch processing, with backpressure to absorb traffic spikes.
My role
Sole author; personal experiment.
Technology stack
- Java
- Cassandra
- Concurrency
- Docker
- Maven
Key engineering decisions
- Design partition keys around time-series telemetry retrieval patterns.
- Implement backpressure in worker queues to handle traffic spikes smoothly.
Lessons learned
- Proper data modeling in wide-column stores like Cassandra is paramount before tackling high-volume ingestion.
Related projects
Configuration Management and Work-Order Platform
Backend Engineering, professional work
Spring Boot microservices for tracking configuration changes, validating templates, and orchestrating work-order workflows across network domains.
- Java
- Spring Boot
- MySQL
- Cassandra
- Docker
- Kubernetes
- Maven