HackTheRounds Interview Experiences

Yelp OA HackerRank Questions 2026: Service Dependencies, Load Balancing & More

Complete breakdown of Yelp's 2026 Online Assessment: service interruption cascades, load balancing, active sellers, and duplicate dish detection.

By HackTheRounds Team · 2026-03-18

Overview

Yelp's 2026 OA uses HackerRank with 2 problems in 70 minutes . The questions are practical and domain-relevant — they model real Yelp engineering problems (service health, load balancing, merchant analytics).

Problem 1: Service Interruption Cascade

Difficulty: Medium | Topics: Graph, DFS/BFS

Given a dependency graph between services, determine the status of each service when outages occur. Outages cascade recursively to downstream services.

Status levels: - Operational: service and all sub-services are up - Partial Outage: some sub-services affected - Complete Outage: service itself or all sub-services are down

Approach: Build an adjacency list, then for each outage, run DFS to mark affected services. Classify each service based on how many of its dependencies are affected.

Problem 2: Load Balancing

Difficulty: Medium | Topics: Simulation, Greedy

Route incoming requests to the server with the lowest load-to-worker ratio. Use server ID as tiebreaker.

Approach: For each request, compute load / workers for all servers, pick the minimum. Update the selected server's load. Straightforward simulation — the key is handling floating-point comparison correctly.

Problem 3: Active Sellers

Difficulty: Medium | Topics: Aggregation, Statistics

A merchant is "active" if it appears in at least 2 event types with above-average frequency. Return sorted list of active merchant IDs.

Approach: Compute per-event-type average across all merchants, then for each merchant count how many event types exceed the average. Filter for count = 2.

Problem 4: Unique Dishes (Only Entrees)

Difficulty: Easy-Medium | Topics: Hashing, Sets

Count unique dishes where two dishes with the same ingredients (regardless of order) are considered identical.

Approach: Sort each dish's ingredient list, convert to a tuple/string, add to a set. Return the set size.

Yelp OA Prep Tips

  1. Domain-relevant problems — expect service health, merchant analytics, and restaurant data scenarios
  2. Graph problems are common — practice DFS/BFS with real-world modeling (dependency cascades)
  3. Clean simulation code — load balancing problems test your ability to write correct, readable simulation loops
  4. 70 minutes is sufficient — don't rush, focus on correctness and edge cases

Browse Yelp interview questions — coming soon to HackTheRounds.