Route Requests to the Nearest Healthy Data Center with Capacity
Quick Overview
Route requests to the nearest healthy data center with free capacity, using explicit geometry, deterministic ties, state updates, and atomic reservation considerations.
Route Requests to the Nearest Healthy Data Center with Capacity
Company: Stripe
Role: Software Engineer
Category: Coding & Algorithms
Difficulty: medium
Interview Round: Technical Screen
Implement a router that selects the nearest healthy data center with available capacity for a request location. Support updating data-center state as capacity or health changes.
### Constraints & Assumptions
The source does not specify coordinate geometry or command strings. Accept a supplied distance function and model each center with ID, location, health, and nonnegative available capacity. For this practice interface, successful routing reserves one capacity unit; select the lexicographically smallest ID on equal distance. Return no center if none is eligible. These are explicit assumptions to confirm.
### Clarifying Questions to Ask
- Are coordinates planar or geographic, and which distance formula is required?
- Does routing consume capacity, or only report a candidate?
- What tie-breaker and update semantics do the original commands specify?
```hint Filter before ranking
An unhealthy or full center must not win merely because it is geographically closest.
```
### What a Strong Answer Covers
- State updates and eligibility filtering before distance comparison.
- Deterministic equal-distance behavior and no-candidate handling.
- Capacity reservation and the concurrency boundary.
### Follow-up Questions
- When would a spatial index help, and how does rapidly changing eligibility complicate it?
- How would you atomically reserve capacity under concurrent requests?
Overview: Route requests to the nearest healthy data center with free capacity, using explicit geometry, deterministic ties, state updates, and atomic reservation considerations.
Implement a router that selects the nearest healthy data center with available capacity for a request location. Support updating data-center state as capacity or health changes.
Constraints & Assumptions
The source does not specify coordinate geometry or command strings. Accept a supplied distance function and model each center with ID, location, health, and nonnegative available capacity. For this practice interface, successful routing reserves one capacity unit; select the lexicographically smallest ID on equal distance. Return no center if none is eligible. These are explicit assumptions to confirm.
Clarifying Questions to Ask Guidance
Are coordinates planar or geographic, and which distance formula is required?
Does routing consume capacity, or only report a candidate?
What tie-breaker and update semantics do the original commands specify?
What a Strong Answer Covers Guidance
State updates and eligibility filtering before distance comparison.
Deterministic equal-distance behavior and no-candidate handling.
Capacity reservation and the concurrency boundary.
Follow-up Questions Guidance
When would a spatial index help, and how does rapidly changing eligibility complicate it?
How would you atomically reserve capacity under concurrent requests?