Greetings fellow software engineers! How have you been doing?

Welcome to the 28th episode of Biweekly Engineering - your source for curated software engineering articles! 💪 

Today, as usual, we have two very intriguing articles to read and learn from!

How Zomato Handled 2023 NYE Traffic

The holidays are a time for celebration, but for backend engineers, holidays can also be a time for controlled panic – especially when it comes to scaling for massive traffic spikes. That's why I found this blog post from Zomato a fascinating one.

In this article, the Zomato engineering team dives deep into the challenges they faced handling the surge in orders on New Year's Eve. The team tackles this challenge head-on, detailing their strategies for:

• Traffic prediction and capacity planning: Accurately forecasting demand is critical. The article discusses the tools and techniques they used to anticipate the NYE rush.

• Scalable infrastructure: No single server can handle a system-wide overload. The post explores how Zomato utilizes cloud-based solutions to automatically scale resources during peak periods.

• Performance optimization: Even with the right infrastructure, efficiency matters. Zomato shares optimizations they implemented to ensure smooth performance under heavy load.

Apart from the preparation, the two critical aspect of the effort were:

• Chaos-testing: Zomato took advantage of chaos-testing, a testing mechanism where failures are intentionally injected in systems to test how the system reacts to the chaos.

• Benchmarking: The team also used their in-house benchmarking tool to test the performance across hundreds microservices.

If you're a backend engineer working on a platform that experiences traffic spikes, I highly recommend checking out this blog post. Zomato offers valuable insights and practical strategies that you can adapt to your own systems.

Interesting Pixels: Demystifying Video Processing

The next post we have in this week's episode dives into the complex world of video processing with this article from Skeepers Tech.

As backend engineers, we often work with data, but video throws a whole new curveball. It's not just static information – it's a complex combination of audio and visual streams that require specialized techniques for processing.

The blog post tackles this head-on, offering a beginner-friendly introduction to video processing concepts. It breaks down the core components you need to understand:

• Video codecs: Codecs are the mechanisms that compress and decompress video data for efficient storage and transmission. Using a codec is called encoding. Examples: h264, VP9, AV1.

• Bit rate: A bit rate is the number of bits processed per second when a video is playing. The higher the bit rate, the higher the quality.

• Adaptative bit rate streaming: As the name suggest, adaptative bit rate is the idea of adapting bit rate based on the internet speed of the consumer during streaming the video. This is why you automatically see lower quality frames for a YouTube video when the internet connection degrades!

Another aspect the articles discussed is how to choose a codec. Like everything in life, each codec has its upside and downside, and you need to evaluate each one to understand which codec solves your problem most efficiently. As an example, the article talked about YouTube videos where most viewed videos are encoded using AV1 codec whereas videos with low view count are encoded with h264.

And that’s it for today. I am leaving you with two very interesting articles. In fact, this was the first time I myself learned about video processing basics. It was worth the read definitely.

See you soon! 🕊️