Shulou(Shulou.com)06/01 Report--

This article will explain in detail what are the interview questions about System design, and the content of the article is of high quality, so the editor will share it for you as a reference. I hope you will have a certain understanding of the relevant knowledge after reading this article.

Developers tend to solve SDI problems because they are so open and often require critical thinking that is not practiced in other coding interview challenges.

Although SDI issues change over time, some of them are still popular in interviews with various top companies.

Today, we will explore the 10 most common system design interview questions, the common questions that must be solved in each question, and some tools that can help you accomplish this task.

Tips for any SDI problem

Start each problem by stating what you know: list all the required features of the system, common problems that such systems may encounter, and the traffic the system wants to handle. The listing process allows the interviewer to review your planning skills and correct any possible misunderstandings before starting the solution.

Describe any trade-off: each system design choice is important. At each decision point, list at least the positive and negative effects of the choice.

Ask your interviewer to clarify: most system design questions are deliberately ambiguous. Ask a clarification question and show the interviewer how you view the problem and your understanding of the system requirements.

Discuss emerging technologies: at the end of each question is an overview of how and where the system can benefit from machine learning. This will show that you are ready not only for the current solution, but also for the future solution.

1. Design a global chat service, such as Facebook Messenger or WhatsApp

For this problem, you will design a service that allows users to chat with each other through Internet. A conversation can be an one-on-one conversation or a group chat with many members. Messages contained in the conversation can only be accessed by those messages.

Must-have features:

Messages must be sent and received over the Internet.

The service must support one-on-one and group chat.

Messages should be stored for later viewing.

Users should be able to send pictures and videos as well as text messages.

Messages should be encrypted during transmission.

Messages should be visible with minimal delay.

Frequently asked questions:

What if I send an email without an Internet connection? Do you want to send after the connection is restored?

How to encrypt and decrypt messages without increasing latency?

How do users receive notifications?

Will the message be pulled from the device (if the server is waiting to send a message, the server will periodically prompt the device) or will the message be pushed to the server (the device will prompt the server to send a message)?

Tools to consider:

The database schema is divided into tables: user tables (with user ID and contacts), chat tables (lists of users with chat ID and participating users ID), and message tables (past messages with references to chat ID).

Use WebSocket for a two-way connection between the device and the server.

Use push notifications to notify members, even if they are offline.

two。 Design a ride-sharing service like Uber or Lyft

This question requires you to create a ride-sharing service that matches users with nearby drivers. The user can enter the destination and send its current location, and notify the nearby driver within seconds.

The application then tracks the route between the driver and the user's current location, and then tracks the route from the user's location to the destination.

Must-have features:

The system must track the current location of all users and drivers.

During transportation, users and drivers must receive updated travel information.

Thousands of users must be supported at all stages of the process and scaled up accordingly.

Drivers and users must always connect to the server.

Frequently asked questions:

How do you maintain low latency during peak hours?

How does the driver pair with the user? It would be inefficient to iterate over all drivers to find the Euclid distance.

What happens if the driver or the user loses the connection?

How do you store all cached location data?

Tools to consider:

Use the S2Geometry library to split the location into cells. Only drivers in the same cell as the user are used to calculate the driver distance.

Distributed storage is used to store the locations of all users, and the location data for each user is only about 1Kb.

If the location data is paused, the device continues to report its previous location while waiting to reconnect.

Please leave a buffer zone after the nearest driver has traveled. If they refuse, go to the next driver.

3. Design URL short services, such as TinyURL or bit.ly

This problem requires you to create a program to shorten a long URL, such as TinyURL or bit.ly. These programs take a long URL and generate a new unique short URL. They can also enter a shortened URL and return the original full URL.

Must-have features:

Return a URL shorter than the original URL

The original URL must be stored

The newly generated URL must be able to link to the original URL stored

Shortened URLs should allow redirection

Custom short URL must be supported

Many requests must be supported at once

Frequently asked questions:

What if two users enter the same custom URL?

What if the number of users exceeds expectations?

How does the database adjust the storage space?

Tools to consider:

Use a hash to link the original and new URL

Use REST API load to balance high traffic and handle front-end client communication

Use multithreading to process multiple requests at a time

Use a NoSQL database to store the original URL (there is no relationship between the stored URL)

4. Design a popular social media service, such as Facebook,Twitter or Instagram

To solve this problem, you will design a social media service for 100, 000 users like Instagram. Users should be able to view news feeds with follower posts and suggest new content that users like.

Interviewers usually want to hear you discuss news sources in depth.

Must-have features:

Powerful news sources and recommendation system

Users can post public posts

Other users can comment or post

Must comfortably accommodate many users at a time

The system must be highly available

Frequently asked questions:

Well-known users will have millions of followers. How will they deal with them compared with ordinary users?

How does the system distribute weight by age? Compared with new posts, old posts are less likely to browse.

What is the ratio of node read to nodes in the write set? Is it possible to have more read requests (users are viewing posts) or write requests (users are creating posts)?

How do you increase availability? How to update the system? What happens if the node fails?

How do you effectively store posts and images?

Tools to consider:

Use rolling updates and replica nodes to maximize availability.

Use a trained machine learning algorithm to recommend posts.

Create a database schema that stores celebrities and users respectively.

Use the social graph to further track the following habits

5. Design a social network and message board service, such as Quora,Reddit or HackerNews

For this problem, you will design a system similar to a forum in which users can post questions and links.

Other users can view and comment on questions. Questions have tags that represent their topics, and users can follow them to view questions about a particular topic. Users have a news source that highlights frequently asked questions from tags and related topics they follow.

Must-have features:

Users must be able to create public posts and apply tags

Posts must be sorted by tag

Other users must be able to post comments in real time.

The database must store data on each post (views, likes, etc.)

News sources must display posts from posts that follow tags as well as other tags that users want.

Must support high-traffic viewers and new posts.

Frequently asked questions:

Does our product only need to run on the network?

Where are the images / links uploaded by users stored?

How will the system determine the relevant labels? How many posts from unfollowed tags are shown in the feed?

How do I distribute posts on the server network?

Tools to consider:

Use SQL database to map relational data (users have posts, posts have comments / likes, categories have related posts, etc.)

The use of multithreading and load balancer layers can help support higher traffic.

Use fragmentation to destroy the system. Consider slicing by category to store posts with the same tags on a computer.

Use machine learning and natural language processing to find the correlation between tags

6. Design global file storage and sharing services, such as Dropbox,Google cloud hard drives or Google albums

For this problem, you will create a synchronous cross-platform storage system, such as Dropbox. Users can store files and photos and access them from other devices.

Must-have features:

Users should be able to save / delete / update / share files over the network

Older versions of the document should be saved for rollback

File updates should be synchronized on multiple devices

Frequently asked questions:

Where are the files stored?

How do you handle updates? Do you want to re-upload the entire file again?

Does the mini update require a complete file update?

How does the system handle two users updating documents at the same time?

Tools to consider:

Use chunks to divide the file into multiple parts. The update only re-uploads that part, not the entire file.

Use cloud storage such as Amazon S3 to handle internal databases.

Keep the client checking with the server to ensure that concurrent updates are applied.

7. Design global video streaming services, such as YouTube or Netflix

This question requires you to create an online video streaming service, such as YouTube. The service will store and transmit hundreds of PB of video data. It must also store statistics (views, likes, views, etc.) and allow users to comment.

Your solution must be scalable to support thousands of concurrent users.

Must-have features:

Videos should be uploaded over the Internet.

Users should receive an uninterrupted stream through the Internet

Video statistics should be stored and accessible to each video.

Comments must be saved and displayed with the video to other netizens for comments

Should support high traffic for thousands of users

Frequently asked questions:

How will your service ensure a smooth video stream on all kinds of Internet quality?

How does your service respond to a sudden drop in flow speed (buffering, quality degradation, etc.)?

How is the video stored?

Tools to consider:

Use cloud technology to store and transmit video data.

Use machine learning to suggest new video content.

Prevent stutter caused by inconsistent connection. Instead of entering the data, the user looked at the data a moment ago.

8. Design API rate limiters for websites such as Firebase or Github

For this problem, you will create an API rate limiter to limit the number of API calls the service can receive within a given period of time to avoid overload.

Interviewers can make requests of all sizes, from one machine to the entire distributed network.

Must-have features:

The device is limited to 10 requests per hour

If their request is blocked, the limiter must notify the user.

Traffic appropriate to its size must be handled.

Frequently asked questions:

How does your system measure hourly requests? If a user makes 10 requests at 1:20 and then another 10 requests at 2:10, they make 20 requests in the same 1-hour window despite the change in hours.

How is the design of a distributed system different from that of a local system?

Tools to consider:

Use the sliding time window to avoid resetting every hour.

Save a counter integer instead of the request itself to save space.

9. Design a proximity server similar to Yelp or nearby places / friends

For the last question, you will design a neighboring server to store and report the distance to places such as restaurants. Users can search nearby locations by distance or popularity. The database must store data from 500 million locations around the world, but with low latency.

Must-have features:

Store up to 500 million locations.

The location must be uniquely identified and have appropriate data, such as quality check and service time.

The search must return results with minimal delay.

Users must be able to search results by distance or quality.

Frequently asked questions:

How do you store so much location data?

How do you get quick search results?

How does your system handle different population densities? A rigid latitude / longitude grid will result in a density-based change response.

How do we optimize common search locations?

Tools to consider:

Use relational databases to store location lists and related data.

Use caching to store data for the most popular locations.

Use shards to split data by region.

Search for locations within a dynamic grid. If there are more than 500 locations in a single cell, divide the grid into 4 smaller cells. Repeat until you only need to search for less than 500 locations.

10. Design search engine-related services, such as Type-Ahead

The service will partially complete the search query and display 5 suggestions to complete the query. It should adapt to highly searched content in real time and make recommendations to other users.

For example, the Seahawks will be advised to win the Super Bowl within minutes of the incident.

Must-have features:

The service should match some queries with popular queries.

Minor spelling errors should be corrected, such as "dgo → dog"

The five most likely options should be guessed from the query

The results should be updated when the query is written

Frequently asked questions:

How good is your ability to correct spelling mistakes?

How do I update the selection without causing delay?

How do you determine which query is most likely to be completed? Does it adapt to the user's search?

What if the user types quickly? Will the recommendation be displayed only after it is completed?

Tools to consider:

Natural language processing machine learning algorithms are used to predict the next character.

Use Markov chain to rank the query probability with the highest ranking.

Update the ML algorithm on an hourly or daily basis rather than in real time to reduce the burden.

What are the System design interview questions to share here, I hope the above content can be of some help to you, can learn more knowledge. If you think the article is good, you can share it for more people to see.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.