The Bitrock engineering team had the opportunity to attend the DevWorld Conference 2024 in the beautiful city of Amsterdam at the end of February. Throughout the event, they had the opportunity to meet with numerous companies at the booths and discuss the present and future of the IT world. They enjoyed a lot of different talks covering many aspects, from discussions on WebAssembly (WASM) and WebAssembly System Interface (WASI) to new interesting ways to embrace and focus on the concept of Green IT, the conference provided attendees with valuable insights and knowledge. This is a summary of some of the talks they found most interesting.
A greener, Cost-Effective Cloud with Serverless WebAssembly
WebAssembly (WASM) has gained considerable attention in the development community for its potential to revolutionize web application development.
At the conference, one of the sessions highlighted the benefits of WASM and its companion standard, WebAssembly System Interface (WASI). WASM offers developers an alternative to traditional JavaScript by allowing them to compile code written in languages such as C, C++ or Rust into a binary format that can be executed in Web browsers. This approach opens up new possibilities for improving the performance and resource utilization of web applications compared to standard JavaScript applications. With WASI, it’s possible to provide access to several operating system-like features, including files and file systems, Berkeley sockets, clocks, and random numbers, which we will enable with WASM outside of the original browser ecosystem.
During the session, there was a fascinating comparison between running a WASI application and running the same application using Docker containers. Data presented at the conference showed various metrics such as startup time, performance and resource usage, and highlighted how WASM performs very similarly to native code when compared to running applications in a Docker container. This comparison highlights the benefits and trade-offs of each approach, providing valuable insights for developers considering the adoption of WASM/WASI in their projects.
Event-driven autoscaling for Serverless Java
Another interesting topic discussed at the conference was the autoscaling of Java serverless applications on Kubernetes. The session introduced attendees to concepts such as Horizontal Pod Autoscaler (HPA), and highlighted the challenges and solutions for dynamically scaling Java applications in a Kubernetes environment.
One of the key highlights was the introduction of KEDA (Kubernetes-based Event-Driven Autoscaling), which provides a robust solution for autoscaling applications based on custom metrics. KEDA provides support for a wide range of external metrics adapters, including Prometheus and Kafka, allowing developers to scale applications based on real-time data from multiple sources simultaneously.
In addition, the session introduced Knative, a Kubernetes-based platform for building and deploying serverless applications. Knative simplifies the deployment and management of serverless workloads on Kubernetes, with features such as scale-to-zero and request-driven scaling for efficient resource utilization.
From 0 to 15 million users: a tech journey on how to scale a mobile app across countries
In this session, representatives from ING shared their experiences of scaling mobile applications across multiple countries and teams. Key highlights from the session included:
- Organizing Projects around Contributions: The teams at ING organized projects around contributions, with each team owning specific modules of the app. They adopted a single native shared code repository and followed trunk-based development practices.
- Product-oriented Organization: ING emphasised organization around the product rather than autonomy, with micro-teams responsible for coding, testing and merging changes. Platform teams took responsibility for post-release operations in production, allowing temporary teams to address specific issues and then disband.
- Modular Tech Platform: ING aimed to be a modular technology platform, using a unified design system and integrating services from other departments, such as security and database services.
- Standardization for Global Customization: Modules were standardized for global customization, allowing optional features to be enabled based on country-specific business requirements.
- Local Configuration Enablement: Local configuration was enabled through embedded web widgets and standalone deployments, allowing for country-specific customisation.
- One Application, One Repository: ING adopted a one app = one repository approach, further emphasising trunk-based development practices.
How LLM and Generative AI Really Work
AI is a very hot topic in the IT world and of course at DevWorld they found a lot of talk about it. One of these sessions was dedicated to exploring Google Gemini, a multimodal AI model designed to handle complex tasks such as conversational AI and natural language understanding, and Gemma, a family of lightweight open models.
Google Gemini provides a powerful framework for building AI applications that require contextual understanding across multiple modalities, including text, images and audio. Developed by Google DeepMind and other teams across Google, Gemma represents the latest advances in AI technology, providing developers and researchers with lightweight yet state-of-the-art models, optimised for performance across multiple hardware platforms, and focused on security and reliability.
Attendees learned about the different versions of Google Gemini and its potential use cases, including automating tasks such as drive-through ordering at restaurants.
How to cut through the noise and look at the physics behind cloud sustainability
Last but not least, a very interesting talk on Green IT, LeafCloud took the stage to present their pioneering approach to cloud computing and sustainability. By installing data centres in residential buildings and using the heat generated by the servers to heat the homes and provide hot water, LeafCloud is redefining the way we think about cloud services and environmental responsibility. LeafCloud’s Chief Sustainability Officer shared insights into the company’s commitment to sustainability and explained how their unique infrastructure design maximises energy efficiency while delivering reliable cloud services. By integrating data centres into residential buildings, LeafCloud minimises environmental impact and improves resource utilisation, creating a more sustainable future for cloud computing.
LeafCloud’s innovative approach not only reduces energy waste, but also contributes to the creation of more resilient and efficient urban ecosystems, demonstrating the potential for technology to drive positive change.
Conclusions
In conclusion, the DevWorld Conference 2024 offered a glimpse into the exciting future of web development with WASM/WASI, serverless architectures, and powerful AI tools like Google Gemini and Gemma. Additionally, the conference highlighted the growing importance of Green IT with LeafCloud’s groundbreaking approach to sustainable cloud computing. By attending this event, the Bitrock engineering team gained valuable insights that will undoubtedly influence their future projects.
Main Author: Simone Esposito, Team Leader and Software Engineer @ Bitrock
