Stateful Systems in Production · a field guide
Stateless is easy.
State is where production gets real.
You can delete and recreate a web pod without anyone noticing. You can't do that to a database, a Kafka broker, or a Redis primary. This course is a field guide to those systems. Every unit starts with a plain mental model, opens up the internals, runs a hands-on lab, lists the configuration knobs that matter, and ends with failure drills: the outages each system is known for, so you recognise them before they page you.
Where each system sits in ShopLite
The guide
Part 0
Foundations of State
Why databases, queues, and caches are harder to run than web servers, and the ideas every one of them is built on.
Part 1
Databases in Production: PostgreSQL
What a DevOps or platform engineer must know about the database behind almost every app: architecture, connections, performance, replication, failover, and safe schema changes.
Part 2
Kafka and Messaging
Event streaming with Kafka from first topic to production operations, and when a classic queue (RabbitMQ, SQS) is the better tool.
Part 3
Redis and Caching
In-memory data structures, caching patterns and their failure modes, memory and eviction, persistence, replication, Sentinel, Cluster, pub/sub, and distributed locks.
Part 4
Search and Log Analytics: Elasticsearch, OpenSearch, ELK
How search engines index text, shards and replicas, cluster health, index lifecycle management, and building a log pipeline with Fluent Bit, Logstash, and Kibana/OpenSearch Dashboards.
Part 5
Performance Engineering
Load testing with k6, JMeter, and Gatling, reading the results correctly, finding the real bottleneck with USE/RED and profilers, and planning capacity with Little's Law.