eBay Software Engineer Interview Experience — Three Onsite Rounds With Two System Designs

Round 1

Given an array heights, where each element represents the height of a vertical line. Choose two lines to act as the walls of a container. Return the maximum amount of water the container can hold (max area).

Given an array of integers temps representing daily temperatures, write a function to calculate, for each day, how many days you'd have to wait until a warmer temperature. The function should return an array answer where answer[i] is the number of days after day i you'd need to wait for a warmer temperature. If there's no future day with a warmer temperature, set answer[i] = 0. The interviewer asked me to walk through an example.

Round 2

Design an app called LunchDrop, mainly for company employees to order food at the workplace. The app has a list of restaurants and their menus, and the available restaurants rotate daily. Users need to be able to:

1. Log into the app and see which restaurants are available that day.
2. Browse each restaurant's menu.
3. Choose food and customize it based on available options (like tacos, vegetables, toppings, sauces, etc).
4. Submit the customized order.

Follow-up questions and discussion on the LunchDrop design:

1. If there's a technical issue at order time, or the menu changes so that an ordered item becomes unavailable, how should the system handle that?
2. What type of database should the menu database use?
3. How do we get updates from real-time data? How does that fit into this system design?
4. Restaurant data also gets updated — where in this system could we use caching?

Behavioral Questions

The interviewer also asked:

1. Why eBay?
2. Why are you leaving your current company?

Round 3: HM Round (Design Round)

Design a dashboard that shows eBay's Top 10 items and Top 10 categories, ranked by sales volume (units sold).

Follow-up questions and discussion on this dashboard design:

1. Explain the item quantity aggregator and category aggregator in detail.
2. How does the system determine the current count for each aggregator?
3. Given eBay's high traffic, how often do we need to write updates to the state store? What are the pros and cons of writing more frequently vs. less frequently?
4. How do we retrieve the final aggregated results from RocksDB?
5. If each partition has its own RocksDB instance, how does the item aggregator maintain and process aggregated data across all partitions?
6. When do we trigger the process of merging data across all partition-level database instances?
7. Suppose there are 1 million events, and each event triggers an update. Does that mean 1 million events get published to the intermediate Kafka topic? If not, how do we optimize this?

Behavioral Questions

The interviewer also asked these situational questions:

1. Have you ever had a conflict with a teammate? How did you resolve it?
2. Why eBay?
3. Have you ever had a conflict with your manager? How did you resolve it?