Hello Reader,

In today's post, let's look at a correct but average answer and a great answer that gets you hired to common cloud interview questions. This should give you pointers on how to improve your interview answers. Also Lambda turned 10 years, should you still learn it? We got a packed newsletter, let's get started.

Interview question - Can you tell me about a microservice you have designed on AWS?​
​

Common but average answer - I will use ALB with EC2 with Auto Scaling Group, ALB with EKS, or API Gateway with Lambda.

Why is it average - What the interviewer is looking for is how the microservice design works. Even a monolith can be run using ALB with EC2 with ASG, ALB with EKS etc. Just saying those doesn’t mean it’s a microservice design. You need to highlight microservice characteristics in the design.

Sample good answer - There are many ways to implement a microservice, but the most common pattern is the one below, with Application Load Balancer (ALB).

Then explain the core concepts:

• Separate target groups handle each microservice. Each target group handles different functions of the application.
• Each functionality can be backed by different compute options. For example, /browse is being served by EC2 + Auto Scaling Group, /buy is being served by AWS Lambda, and Kubernetes are serving others.
• The ALB is associated with a DNS system such as Amazon Route53, so that users can access the ALB using a nice domain name instead of the cryptic default URL ALB provides you
• Not only these three microservices are handling different functionalities, they can scale independently and be coded in different languages which are strengths of using microservices. For example, /browse is written in python, /buy in Node JS, and /* in Go. This property is called polyglot

If you get this question in your interview, make sure to knock it out of the park! Now, let's introspect Lambda's Big Birthday.

Lambda turned 10 years 🍰

When AWS Lambda was announced in 2014, it created lot of buzz. But where are we now, after 10 years. As always, we take a pragmatic look to maximize your preparation for SA interview and real-world projects.

Just for reference, I have been working with Serverless for a very long time, since 2016. I deployed a Serverless application in production at a large enterprise company, Verizon. Whatever I am sharing is based on my actual experience and from working with actual customers, and not just theory crafting.

• Lambda is used heavily for Event Driven - Lambda is the best glue (not to be confused with AWS Glue service!) between AWS services. For example - want to run business logic when a file is put into a S3 bucket? Want to process a message from SQS and insert into a Database? Want to react to an event like EC2 start? Lambda is the answer. For this reason, Lambda is heavily used in Event Driven Architecture (EDA). Popular EDA services include SQS, SNS, EventBridge, S3, along with Lambda.

• Lambda never got rid of "Cold Start" label - When Lambda came out, cold start was indeed an issue. A Lambda cold start refers to the latency or delay that occurs when an AWS Lambda function is invoked for the first time (or after a period of inactivity) and the underlying infrastructure needs to initialize before executing the function. This adds a bit of latency. However, Lambda has added various measures to combat this - Provisioned Concurrency, Snap Start, different techniques like moving static code to the global section, identifying cold start code via X-Ray etc. Even in cases where the cold start is zero or minimal, unfortunately, even now when I go to customers, they always ask about the cold start. It's like one of those labels which is hard to shake off. This perception definitely reduced Lambda adoption. Pro tip - For interviews, ensure to study cold start solutions.

• Study API Gateway with Lambda - Lambda is the crown jewel of Serverless. However Serverless is a principle and not a singular service. Serverless broadly implies these four rules - a/ no servers to manage, either physical or virtual, b/ scales automatically, c/ pay as you go, and d/ inherently highly available. One mistake candidates often do is they study Lambda, but ignore the other major Serverless service - API Gateway. Amazon API Gateway is adopted independent of Lambda, and brings many powerful features such as rate limiting, payload transformation, easy Authn/z, caching (yes, caching!),swagger support, and tada - integration with numerous AWS services along with Lambda. For this reason, it's a popular service in interviews, and real world projects.

• Elephant in the room - Kubernetes - Almost all customer meetings I handle where customer is modernizing their cloud app, they'd always ask about Lambda and...Kubernetes, or containers. There is no denying that GCP and Azure are growing as well. Though most companies select one cloud as a primary provider, and that's the fastest and cheapest way to go, customers want to keep the options open. For that reason, given the portability of containers, Kubernetes has shot up in demand. The good news is that large companies like Cap One, FINRA, and many others have adopted Lambda. Ideally, you should know both Lambda and Kubernetes, but obviously, you are thinking, what if you have to pick one? Because Kubernetes is cloud agnostic, there are more jobs on K8s. For that reason, I'd say, if you want to learn just one, learn Kubernetes (Listen, I told ya I don't BS and always give you the pragmatic answer ;)).

• Popular Lambda interview topics - Lambda is still important for interviews because companies always evaluate and adopt it for modern applications. This brings us to popular Lambda interview topics. Lambda properties, how is Lambda charged, cost/performance optimization, securing Lambda, Lambda with API gateway, Lambda cold start, tracing, Lambda vs EC2 vs Kubernetes, EDA, Lambda in and outside VPCs are important topics.

I want to end this newsletter with some fun memories. After I deployed a Lambda application in Prod in 2017, my talk was selected for a conference in San Francisco. This was my very first conference talk!

I became so passionate about Lambda, that I created a Serverless course on 2018, and also put a Lambda light on my home office wall! It's still shines brightly today and present on my videos and customer calls.

Here's to next 10 years of AWS Lambda 🙌🚀

If you have found this newsletter helpful, and want to support me 🙏:

Checkout my bestselling courses on AWS, System Design, Kubernetes, DevOps, and more: Max discounted links​

AWS SA Bootcamp with Live Classes, Mock Interviews, Hands-On, Resume Improvement and more: https://www.sabootcamp.com/​

Keep learning and keep rocking 🚀,

Raj