Have a System Design Interview coming up? Breeze through the basics and get interview-ready in hours through a selection of hand-picked practice problems.

sys-design.tar.gz

basic

System design interviews play a vital role in hiring software engineers and managers. These interviews test the ability of the candidate to solve problems under different constraints. A candidate’s performance in a system design interview is a significant factor in deciding the hire/no-hire decisions and the level assigned to the incoming candidate.

In this course, you’ll go through a miniature form of our primary course on system design: Grokking Modern System Design Interview for Engineers & Managers. This crash course aims to give you an overall gist of the system designing process using short, easy-to-digest problems. Each design problem of the course can be completed in about 15 minutes, giving you the flexibility to include them in your jam-packed schedules.

System Design Interview Prep Crash Course

# What is Yelp?

**Yelp** is a one-stop platform for consumers to discover, connect, and transact with local businesses. Users can easily join a waitlist, make a reservation, schedule an appointment, or purchase goods. Yelp also provides information, photos, and reviews about local businesses.

# Requirements

Let’s identify our system's requirements.

## Functional requirements

The functional requirements of our system are below:

* **User accounts:** Users will have accounts where they’re able to perform different functionalities, like log in, log out, add, delete, and update.

  > **Note:** There can be two types of users: business owners who can add their places on the platform and other users who can search, view, and rate a location.

* **Search:** The users should be able to search for nearby places or places of interest based on their GPS location (longitude/latitude) and/or their name.

* **Feedback:** The users should be able to add a review about a place. The review can consist of images, text, and a rating.

## Design 

Now, we’ll discuss the individual building blocks and components used in the design of Yelp and how they work together to complete various functional requirements.

### Components

These are our system's components:

* **Segments producer:** This component is responsible for communicating with third-party world map data services (for example, Google Maps). It takes that data and divides the world into smaller regions called segments. The segment producer helps us narrow down the number of places to be searched.

* **QuadTree servers:** These are a set of servers that have trees that contain the places in the segments. A QuadTree server finds a list of places based on the given radius and the user’s provided location and returns that list to the user. This component mainly aids the search functionality.

* **Aggregators:** The QuadTrees accumulate all the places and send them to the aggregators. Then, the aggregators aggregate the results and return the search result to the user.

* **Read servers:** We use a set of read servers that we use to handle all the read requests. Since we have more read requests, it’s efficient to separate these requests from the write requests. Each read server directs the search requests to the QuadTrees’ servers and returns the results to the user.

* **Write server:** We use a set of write servers to handle all the write requests. Each write server handles the write requests of the user and updates the storage accordingly. Examples of write requests include adding a place, writing a comment, rating a place, and so on.

* **Storage:** We’ll use two types of storage to fulfill our diverse needs.

   * **SQL database:** Our system will have different tables, like “Users,” “Place,” “Reviews,” “Photos,” and others, as described below. The data in these tables is inherently relational and structured. We need to perform queries, like places a user visited, reviews they added, or view all the reviews of a specific place. It’s easy to perform such queries in a SQL-based database. We also want all users to have a consistent view of the data, and SQL-based databases are better suited for such use cases. We’ll use reliable and scalable databases, as is discussed in the lesson [Databases in Distributed Systems](https://www.educative.io/courses/system-design-interview-prep-crash-course/databases-in-distributed-systems).

  * **Key-value stores:** We’ll need to efficiently fetch the places in a segment. For that, we store the list of places against a segment ID in a key-value store to minimize searching time. We also save the QuadTree information in the key-value store by storing the QuadTree data against a unique ID.

* **Load balancer:** A load balancer uniformly distributes users’ incoming requests to all the servers.

Learn how to design a proximity service like Yelp.

Introduction

Elementary Design Problems

Advanced Design Problems

Concluding Remarks

Design a Proximity Service/Yelp

What is Yelp?

Requirements

Functional requirements

Non-functional requirements

Building blocks we will use