How do I approach System Design interview questions?

System Design questions require understanding of core concepts and practice. PracHub provides solutions with explanations to help you master system design interviews.

What difficulty level is this interview question?

This is a hard difficulty System Design question, commonly asked during Technical Screen rounds at Snowflake.

What role is this question designed for?

This question is commonly asked for Software Engineer candidates at Snowflake during technical interviews.

Design under vague distributed requirements | Snowflake Interview Question

Quick Overview

This question evaluates expertise in distributed system architecture, metadata and schema management, transactional DDL semantics, consistency and replication strategies, API and data-model design, multi-tenant isolation, observability, security, and capacity planning.

System Design Prompt: Distributed Metadata Catalog and Schema Registry

Context

Design a multi-tenant distributed Metadata Catalog and Schema Registry for an analytics platform. The service manages databases, schemas, tables, columns, views, roles, and grants. It must support high read throughput, transactional DDL updates, and change notifications to downstream systems (e.g., query planner, caches).

Begin by clarifying requirements (functional and non-functional), then design the system end to end.

Tasks

Clarify requirements
- Enumerate core use cases and access patterns.
- Define SLAs/SLOs (latency, availability, durability, multi-region needs, data retention).
- Specify consistency expectations and isolation levels.
- Identify multi-tenancy constraints and per-tenant limits.
External APIs
- Propose REST/gRPC endpoints for CRUD on entities, conditional updates, transactions, and change-subscription.
- Show request/response shapes at a high level, including idempotency and versioning.
Data model
- Define entities and relationships (normalized vs denormalized).
- Describe versioning, soft deletes, and change-log design.
Sharding and replication strategy
- Explain partitioning keys and how to minimize cross-shard transactions.
- Choose replication factor, placement (AZ/region), and read/write topology.
Consistency model
- Justify strong vs eventual consistency per operation type.
- Describe transaction approach (single-shard vs multi-shard).
Read/write paths
- Describe end-to-end request flow for reads and writes, including cache interaction and index maintenance.
Failure handling
- Timeouts, retries, idempotency, partial-failure handling, dead letter policies.
- Leader election and membership changes.
Load management
- Backpressure/admission control, hot-key mitigation, rate limiting.
- Caching strategy (L1/L2, invalidation/signaling).
Schema evolution
- Compatibility rules, online migrations, and rolling upgrades.
Observability
- Key metrics, logs, tracing, and alerting tied to SLOs/error budgets.
Security and compliance
- Authentication, authorization (RBAC), encryption, audit logging, tenant isolation.
Capacity planning
- Provide a simple sizing model with example numbers (QPS, storage, replication overhead, headroom).
High-level architecture and trade-offs
- Present the component diagram verbally and discuss major trade-offs (latency vs availability, complexity vs robustness, cost vs performance).

Quick Overview

Context

Begin by clarifying requirements (functional and non-functional), then design the system end to end.

Tasks

Clarify requirements

Enumerate core use cases and access patterns.
Define SLAs/SLOs (latency, availability, durability, multi-region needs, data retention).
Specify consistency expectations and isolation levels.
Identify multi-tenancy constraints and per-tenant limits.

External APIs

Propose REST/gRPC endpoints for CRUD on entities, conditional updates, transactions, and change-subscription.
Show request/response shapes at a high level, including idempotency and versioning.

Data model

Define entities and relationships (normalized vs denormalized).
Describe versioning, soft deletes, and change-log design.

Sharding and replication strategy

Explain partitioning keys and how to minimize cross-shard transactions.
Choose replication factor, placement (AZ/region), and read/write topology.

Consistency model

Justify strong vs eventual consistency per operation type.
Describe transaction approach (single-shard vs multi-shard).

Read/write paths

Describe end-to-end request flow for reads and writes, including cache interaction and index maintenance.

Failure handling

Timeouts, retries, idempotency, partial-failure handling, dead letter policies.
Leader election and membership changes.

Load management

Backpressure/admission control, hot-key mitigation, rate limiting.
Caching strategy (L1/L2, invalidation/signaling).

Schema evolution

Compatibility rules, online migrations, and rolling upgrades.

Observability

Key metrics, logs, tracing, and alerting tied to SLOs/error budgets.

Security and compliance

Authentication, authorization (RBAC), encryption, audit logging, tenant isolation.

Capacity planning

Provide a simple sizing model with example numbers (QPS, storage, replication overhead, headroom).

High-level architecture and trade-offs

Present the component diagram verbally and discuss major trade-offs (latency vs availability, complexity vs robustness, cost vs performance).

Design under vague distributed requirements

Quick Overview

System Design Prompt: Distributed Metadata Catalog and Schema Registry

Context

Tasks

Solution

Comments (0)

Design under vague distributed requirements

Quick Overview

System Design Prompt: Distributed Metadata Catalog and Schema Registry

Context

Tasks

Solution

Comments (0)