System Design Of Music Streaming App: Everything You Need to Know

We all love streaming music online. Music streaming apps like Spotify have become an integral part of our day-to-day life. But, have you ever wondered how your favorite music streaming apps work? Or, have you ever thought of making your own music streaming app, just like Spotify?

If yes, then you are on the right page! In this blog, we will dive deep into the system design and technical workflow of a music streaming app, and explain how they function so smoothly. So, let’s get started!

System Design For A Music Streaming App

To begin with, let’s think about what elements you see when you’re using any music streaming app. Let’s take the example of Spotify. So in Spotify, you will find the following elements:

1. Songs/Music
2. Playlists
3. Users
4. Artists
5. Podcasts

Now let’s look into the key functionalities of a music streaming app. So, any music streaming app performs two major functions: 

1. Finding and Playing Music [ Search → Play (Browse)]
2. Playing Music [ Desired music on frontend → Just Play ]

Though there can be other functions too, these two are the key or core functions. 

So, to begin system designing for such a music streaming app, we will have to make some high-level assumptions, related to some key metrics. Here we will basically be assuming the amount of data the app will be dealing with. 

Here’s a basic set of assumed key metrics calculation:

• No. Of Users = 1 Billion Users
• No. Of Songs = 100 million songs
• MP3 Audio = 5 MB (typical size)
• Total Audio = 500 TB = 0.5 PB.
• Replication Factor [If data gets lost from database / OR  keep it concerning geological metrics] = 3 x 0.5 = 1.5 PB.
• Metadata = 100B per song = 10GB song metadata
• 1 KB per user = 1 TB

Now that we have done the calculations of basic metrics, let me introduce you to the basic components of any audio streaming app:

1. Frontend (App) UI
2. Load Balancer
3. Web Servers
4. Database        
   1. Blog DB → Amazon S3  (Song Audio DB)
   2. Metadata DB
5. CDN

The basic system design for my music streaming app will look something like this:

Now let’s see how this system will perform the key functions of a music streaming app that we discussed above. 

Browsing Music

A user will search for particular songs that they like. Technically, this step is known as “Run Query”. The music streaming app will receive their query and send their request to find music to a web server. 

Now, this translated query is sent to RDB. The RDB then reverts back with, let’s assume, a set of 100 songs as per your typed keyword. 

So, the user can now look at the 100 songs appearing on their screen. This step technically involves displaying a list of songs received from the RDB.  

After browsing through the songs, let’s assume that the user has decided to play a song. They will “Click” on the selected song, initiating the next functionality of “Playing”.                                                                      

Playing Music

When a user clicks on that particular song that he selected, the “Play” functionality is initiated. This is how the system works:

• The app will send the request to the Webserver.
• The audio link will be fetched from RDBMS.
• The web server will fetch that particular song from  BlobDB.
• The web server will return the song to the app.
• The app audio player will play the song. 

A Possible Bottleneck Situation, And Way to Solve It

Imagine this, the popular K-pop band BTS has just released their new and much-awaited song. Fans were eagerly waiting for it, and the news of its release spread like wildfire. As a result, the requests on your music streaming app suddenly increased a thousandfold. There are 1000 requests/sec for that particular song. 

What will happen now? This sudden spike in requests can overload the database, as the system is receiving tons of requests for the same song simultaneously. 

So, what’s the solution for it? – A CDN (Content Delivery Network)! 

A CDN is basically a kind of cache memory. Please note that caching is very important here so get out of this bottleneck of hot and trending songs. 

So, whenever the web server gets requests for the same songs more frequently, the web server redirects it to CDN instead of sending that request to the database. 

At this point, it’s important to talk about the “Load Balancers”. Load Balancers’ job is to prevent web servers from overloading so that they can provide uninterrupted service to end users. When there are a lot of requests coming to the web server, the Load balancers redirect the request to the nearest web server.  

To Conclude

The system design of a music streaming app like Spotify incorporates various essential components to ensure seamless browsing and playback experiences for millions of users globally. 

Through a combination of frontend UI, load balancers, web servers, databases for audio and metadata storage, and content delivery networks (CDNs), the architecture efficiently manages user requests and optimizes data retrieval. 

By distributing load, caching frequently accessed content, and employing geographical routing, the system maintains high performance and reliability, enabling users to enjoy their favorite music effortlessly.

If you are looking to create an amazing music streaming app like Spotify, but are not so confident in coding, Muvi One is here to help you out! You can create an amazing feature-packed and own-branded music streaming app without writing a single line of code, using Muvi One! 

You get everything out of the box, including CDN, DRM, monetization options, player, and more! You can customize the look and feel of your platform, and develop applications for 16+ OS environments including Android, iOS, Android TV, Roku, and more! Sign up to try Muvi One for FREE for 14 days!