Interactive visualizations for High-Level Design concepts. Built while preparing for HLD interviews — each topic gets its own playable demo so the trade-offs become obvious by clicking, not by reading.
Build a hash ring, add servers, watch keys redistribute.
Toggle virtual nodes from 1 to 200 and see load imbalance collapse.
The canonical demo for understanding why hash % N
fails and why distributed caches use what they use.
Round Robin vs Least Connections vs Power of Two Choices vs Least Response Time. Send requests, simulate a slow server, watch each algorithm react in real time.
Simulate a network partition. Choose CP or AP. See exactly which reads succeed, which fail, and which return stale data — and what that means for your users.
Trace a request from browser to origin. Watch DNS resolution, geo-routing, edge caching, and cache-miss fallback to origin — the path every request takes.
Cache-aside, write-through, write-back, write-around. Simulate cache hits and misses, watch eviction policies in action.
Range-based vs hash-based vs directory-based sharding. See data distribution, hot-shard problems, and rebalancing live.
Leader-follower, multi-leader, leaderless. Tune R, W, N values. See when reads see stale data and how quorum prevents split-brain.
Producers, partitions, consumer groups, offsets. Watch ordering guarantees, rebalancing, and back-pressure unfold visually.
Token bucket, leaky bucket, fixed window, sliding window. Slam the API with traffic and see how each algorithm responds.
Add items, query membership, tune bits and hash functions. Watch false-positive rate change as the filter fills up.
Distributed transactions without 2PC. Trigger a failure mid-saga, watch compensating actions run. Outbox pattern keeps events durable.
Closed → open → half-open. Inject failures into a downstream service and watch the breaker protect the system from cascade.
Duplicate webhooks, network failures, retry storms. Watch idempotency keys save the day and exponential backoff calm the chaos.
System design concepts live in trade-offs — and trade-offs are almost impossible to feel from a textbook diagram. Building these interactive demos is how I'm preparing for HLD interviews: once you can move a slider and watch a metric collapse, you understand the concept in your bones.
I'm Uday Teja — backend engineer udayteja.com · github · leetcode