Real-Time Analytics Dashboard

Interview-Ready Approach

1) Clarify Scope and SLOs

• •Problem statement: Design a real-time analytics pipeline that ingests clickstream events and powers live dashboards with < 5 s latency.
• •Design for a peak load target around 750 RPS (including burst headroom).
• •Peak events/second: ~100,000
• •Data freshness: < 5 seconds
• •Retention period: 90 days
• •Total events stored: ~500,000,000,000
• •Dashboard query latency: < 2 seconds

2) Capacity Planning Method

• •Convert traffic and growth constraints into request rate, storage growth, and concurrency budgets.
• •Keep at least 2-3x safety margin per tier (ingress, compute, storage, async workers).
• •Reserve explicit latency budgets per hop so p95 can be defended in review.

3) Architecture Decisions

• •Analytics: Maintain separate OLTP and analytics paths; stream events into a warehouse/time-series layer.
• •Databases: Define a clear system-of-record and design read/write paths separately before adding optimizations.
• •Message Queues: Move non-blocking and retry-heavy work to async consumers with explicit retry and DLQ policies.
• •Caching: Put cache on hot read paths first and pick cache-aside or write-through explicitly.
• •WebSockets: Use persistent connection gateways and decouple fanout via pub/sub or queues.

4) Reliability and Failure Strategy

• •Version event schemas and monitor drop/late-event rates.
• •Use strong write constraints (transactions or conditional writes) and explicit backup/restore strategy.
• •Guarantee idempotent consumers and trace every message with correlation IDs.
• •Bound staleness with TTL + invalidation hooks for critical entities.
• •Track connection churn, backpressure, and session resumption behavior.

5) Validation Plan

• •Run one peak-load test, one dependency-degradation test, and one failover test.
• •Verify idempotency for all retried writes and async consumers.
• •Track user-facing SLOs first: p95 latency, error rate, and successful throughput.

6) Trade-offs to Call Out in Interviews

• •Analytics: Analytics pipeline unlocks insights, but adds eventual consistency and governance overhead.
• •Databases: SQL gives stronger transactional guarantees; NoSQL often gives better write scaling and flexibility.
• •Message Queues: Async pipelines absorb spikes well, but increase eventual-consistency complexity.
• •Caching: Higher hit rate cuts latency/cost, but stale data and invalidation bugs become primary risks.
• •WebSockets: WebSockets reduce interaction latency but complicate scaling and state management.

Practical Notes

• •Two paths: a real-time path (Kafka → Flink/stream processor → pre-computed counters in Redis) for live dashboards, and a batch path (Kafka → data lake → columnar store) for historical queries.
• •Use a columnar database (ClickHouse, Druid, or Apache Pinot) for fast analytical queries over billions of rows.
• •Pre-aggregate common metrics (active users, event counts) in materialized views to avoid scanning raw events on every dashboard load.