Skip to main content

7-step process

note

These notes are for me to revise all my concepts. I do not claim this to be my own material. The main reference for the entirety of this doc-note is Grokking the System Design Interview.

1. Requirements​

The first step to design any product/system is to understand the requirements. There is no ONE correct answer. Questions to be asked:

  • What features are part of the system
  • Now drill into some finer details for each feature
  • Think of some branching scenarios (alternate paths the user may choose)
  • Confirm whether only the backend is being designed or the frontend too
  • Ask for functionalities that are common across systems (eg. notifications, search)
  • Also question about what type of data is being handled (media, text, etc.)

2. System interface definition​

To follow up on the gathered requirements, we define the API interface. Defining what we take as input and what we give as the output for all the major functionalities helps us decide whether our requirements are complete.

Functionality Definitions:

functionality1(data_type input1, data_type input2, ...)
functionality2(data_type input1, data_type input2, ...) 
functionality3(data_type input1, data_type input2, ...)

3. System scale definition​

We now need to confirm how many users and how much data our system is expected to handle. Technologies and solutions we use will depend on this particular step.

Questions to be asked:

  • What is the scale of the system? (Open ended question, extract as much info as possible)
  • How much storage does the system need? (Ask explicity based on the various file types being handled)
  • What's the expected network usage and bandwidth? (Load balancing and traffic management needs)

4. Define data model (Entities)​

Data flow is decided by first defining the data models. Define which entities are part of the system and their attributes.

Defintion:

Entity1: ent1id, attr1, attr2, ...
Entity2: ent2id, attr1, attr2, ...
Entity3: ent3id, attr1, attr2, ...

Now determine which DB is the best possible solution: NoSQL or SQL. What kind of block storage should we use to store photos and videos?

5. HLD (Diagram)​

Now draw a diagram with just core functionalities. May include some basic load balancer, and identify all important/core components of the system.

6. Detailed design​

Let interviewer guide on details of particular components/functionalities. Keep taking feedback throughout. Think out loud -- discuss pros and cons, tradeoffs between choice of technologies.

Some examples given in Grokking the System Design Interview (twitter example):

  • Since we will be storing a massive amount of data, how should we partition our data to distribute it to multiple databases? Should we try to store all the data of a user on the same database? What issue could it cause?
  • How will we handle hot users who tweet a lot or follow lots of people?
  • Since users’ timeline will contain the most recent (and relevant) tweets, should we try to store our data in such a way that is optimized for scanning the latest tweets?
  • How much and at which layer should we introduce cache to speed things up?
  • What components need better load balancing?

7. Identify & resolve bottlenecks​

Discuss approaches to mitigate these bottlenecks.

Questions to be asked:

  • Is there any single point of failure in our system? What are we doing to mitigate it?
  • Do we have enough replicas of the data so that if we lose a few servers we can still serve our users?
  • Similarly, do we have enough copies of different services running such that a few failures will not cause total system shutdown?
  • How are we monitoring the performance of our service? Do we get alerts whenever critical components fail or their performance degrades?

Overview​

We have seen a total of 7 systematic steps to tackle any HLD question.

  1. Requirement gathering
  2. Interface definition (API)
  3. System scale definition (user size, storage size, network bandwidth)
  4. Entity definition
  5. HLD (Basic diagram)
  6. Detailed design (Expand diagram)
  7. Identify and mitigate bottlenecks

Thus, an organized and systematic way of tackling any HLD problem shows a structured way of thinking. Stay calm and think in terms of how you would tackle it if you were to really build the product.

Last updated on