Plate I · Infrastructure
The Cloud Topology
A reference topology for the systems I design: traffic in at the edge, events through Kafka, truth in the stores, answers out through the API. Hover or click a node to inspect its role and trade-offs.
SELECT A NODE — hover to trace, click to inspect its role and trade-offs.
WALKTHROUGH
- 1.
Client sends request
- 2.
Traffic enters load balancer
- 3.
API Gateway routes request
- 4.
Microservices read/write cache and database
- 5.
Async work is pushed to message queue
- 6.
Workers consume and process jobs
- 7.
Response returns to client
Designing for failure
Every box on this plate can fail. Circuit breakers at the gateway, idempotent consumers on Kafka, read replicas on Postgres, and replayable events mean no single failure cascades.
Separating writes from reads
Commands go through the API and commit to Postgres. Read models are projected asynchronously by workers into Postgres, Elasticsearch, and caches. Each store is tuned for its query shape.
Cost-aware scaling
Horizontal on stateless layers (API, workers), vertical and replicated on stateful ones (Postgres). Caches soak up the hot path. Object storage holds the long tail at tiny per-GB cost.