Compare core system design tradeoffs with scenarios
Overview
This question evaluates competency in system design and architectural trade-offs across service decomposition, data storage models, communication patterns, API protocols, caching, consistency, partitioning, and load balancing, with attention to scalability, latency, availability, cost, and operability.
BitGo
Aug 12, 2025, 12:00 AM
Software Engineer
Technical Screen
System Design
0
0
System Design Tradeoffs for a Web-Scale Service
Context
You are designing a web-scale, user-facing transactional service. Compare key architectural choices across the following dimensions, and ground your answers with quantitative targets (e.g., 20k QPS, P99 200 ms, 99.99% availability):
- Service decomposition: monolith vs microservices
- Primary data store: relational vs document vs wide-column
- Inter-service communication: synchronous RPC vs event-driven messaging
- API protocol: REST vs gRPC
- Caching strategies: write-through, write-back, write-around
- Consistency models: strong vs eventual
- Data partitioning: range, hash, consistent hashing
- Load balancing: round-robin, least-connections, consistent hashing
Task
For each dimension:
- Provide a concrete scenario and quantitative targets (throughput, latency SLOs, availability, retention, etc.).
-
Justify a choice, and discuss implications for:
- Latency
- Availability and durability
- Cost and scalability
- Operability (developer productivity, observability, change management)
- Failure modes and mitigations
Where helpful, include small calculations (capacity, shard counts, cache hit/miss effects) and brief formulas. Keep each comparison grounded in realistic, web-scale use cases.
Solution
Login required
Show