August 2022 Summaries

Kong Gateway, built on Nginx/OpenResty, faced challenges in efficiently sharing information between its isolated worker processes, prompting a shift from the lua-resty-worker-events library to the newly developed lua-resty-events library. The older library relied heavily on shared memory and polling, which resulted in high overhead and inefficiencies, particularly under conditions of heavy traffic. The new lua-resty-events library, integrated into Kong 3.0, leverages Nginx's native event loop and a publish/subscribe system using UNIX domain sockets to facilitate asynchronous message dispatch without the drawbacks of shared memory, thus improving performance and reducing CPU usage. Benchmark tests demonstrated that lua-resty-events significantly enhances the handling of massive events compared to its predecessor, offering lower latency and more stable CPU consumption. This transition reflects an advancement in Kong Gateway's capability to manage large-scale cloud-native traffic, with ongoing development aimed at further enhancing the library's features and efficiency.

The Kong Champions program is designed to celebrate and support members of the Kong community who demonstrate exceptional dedication to the company's products and open-source projects, fostering a global network of enthusiastic contributors. With 18 Champions worldwide, the program emphasizes passion, inclusivity, and unity by offering free training, direct communication with Kong's teams, and various incentives, such as swag and unique gestures like gorilla adoptions for nature conservation. The program's participants come from diverse backgrounds and locations, sharing their expertise in forums, blogs, and events while feeling a sense of family and support within the community. Kong recognizes the importance of creating meaningful connections and giving back to the world, reinforcing the Champions' role as integral to the company's mission and values.

Kong Mesh 1.9 and Kuma 1.8 have been released with a host of new features and improvements designed to enhance service mesh capabilities, including the introduction of observability for the built-in Gateway with features such as tracing, metrics, and an accompanying Grafana dashboard. The Cross-Mesh Gateway now supports multi-zone functionality, allowing traffic routing between zones where gateways are present, and the CNI has been rewritten to leverage a new transparent-proxy. Additionally, the Mesh Gateway supports TCP traffic and provides options for path rewrite and header modifications, while the retry policy offers greater flexibility, and metrics filtering is introduced to enhance performance. Security is bolstered with support for Projected Service Account Tokens in Kubernetes, and a complete rewrite of the TCP TrafficLog aligns with Envoy's capabilities. These updates aim to simplify multi-tenant and multi-zone use cases, improve production readiness, and offer a more feature-rich and observant gateway experience.

Web3, or Web 3.0, is envisioned as the third iteration of the internet, focusing on decentralization and the minimization of centralized control, utilizing technologies like blockchains and cryptocurrencies. It follows the eras of Web 1.0, characterized by static pages, and Web 2.0, known for its participatory and social web platforms. Advocates of Web3 argue that it offers solutions to issues present in Web 2.0, such as centralized control and data monetization without user benefit, promoting a more transparent and permissionless internet. While Web3 introduces innovative concepts like decentralized payments and smart contracts, it faces challenges in user adoption due to its complexity, particularly for non-technical users. Frontend developers exploring Web3 encounter new foundational elements but can leverage existing skills in HTML, CSS, and JavaScript. The hybrid approach, integrating both Web2 and Web3 technologies, is suggested to enhance user experience and facilitate onboarding. Despite its promising potential, achieving mass adoption of Web3 is hindered by user experience issues, primarily stemming from the steep learning curve and the necessity for users to manage their own private keys.

Decentralization in systems architecture, especially within Web3, encompasses various concepts, including physical, governance, and permissionless decentralization. Physical decentralization involves autonomous nodes that maintain system functionality even if some nodes fail, akin to horizontal scaling in Web2. Governance decentralization, prevalent in Web3, allows multiple unrelated parties to manage systems through a proposal and voting process, often facilitated by Decentralized Autonomous Organizations (DAOs). Permissionless architectures enable anyone to add nodes without central authority approval, provided they meet hardware and software criteria. Blockchain networks like Ethereum, Polygon, Solana, and Polkadot exemplify decentralization by employing consensus algorithms such as Proof of Work (PoW) and Proof of Stake (PoS) to maintain network integrity and resist malicious activity. These networks often use HTTP gateways, which can be public or private, to allow system access, with private gateways offering higher performance for a fee. Decentralization in Web3 aims to eliminate single points of failure, enhance security, and promote collaborative decision-making across a global community.

Chaos Engineering is a proactive approach to enhancing the resilience of distributed systems by intentionally injecting controlled failures and observing the system's responses. The practice originated from Netflix's need to design for failure after a significant outage, leading to the creation of Chaos Monkey, which tests system weaknesses by intentionally disabling instances. This method has since gained popularity, with tools like AWS Fault Injection Simulator and increasing interest in the field. By simulating failures, Chaos Engineering helps teams identify vulnerabilities, improve system documentation, and build reflexive responses to real-world failures, ultimately leading to higher availability and decreased Mean Time To Resolution (MTTR). The blog series emphasizes the importance of Chaos Engineering for any platform, using Kong Gateway as an example to illustrate scenarios like testing hybrid deployments and availability zone outages. Through these controlled experiments, teams can prepare for potential failures, ensuring their systems are robust and minimizing downtime, thus maintaining confidence in the platform's reliability.

As companies increasingly embrace digital transformation, characterized by a surge in APIs and microservices, optimizing data collection and feedback is crucial for enhancing user and developer experiences. Kong Gateway, an adaptable and lightweight API gateway, facilitates this process through its extensive plugin ecosystem, allowing for efficient monitoring and logging across various technology stacks. A specific case explored involved forwarding API logs to Azure Log Analytics using Kong deployed in Azure Kubernetes Services, necessitating third-party plugins for Logstash and FluentD to bridge the gap, as there are no native solutions available for direct log shipping. The process involved configuring these plugins using Helm, setting up sample services, and generating API logs to ensure seamless integration and effective log analysis within Azure Log Analytics. This experimentation highlighted Kong's ability to integrate effortlessly with other solutions, simplifying the complexity of integration and providing valuable insights into microservices behavior, thereby enhancing operational capabilities for businesses.

APIs have become critical components in the digital economy, allowing businesses to interconnect systems, enable transactions, and create new business models. The evolution of APIs has been significantly influenced by cloud computing, with modern APIs facilitating cloud connectivity and supporting polycloud strategies that leverage multiple public cloud platforms. Common API architectures include RESTful APIs, gRPC, and GraphQL, each offering unique advantages for data communication. The rise of microservices, serverless functions, and service meshes has further promoted the development of loosely coupled API designs, enhancing scalability and manageability. In the future, APIs are expected to play a crucial role in Web 3.0, becoming more event-driven to align with decentralized networks and protocols. Additionally, APIs are integral to advancements in artificial intelligence, IoT, and blockchain technologies, driving innovations in fields such as fintech and cybersecurity. With ongoing improvements in API development practices, performance, and security, tools like Kong's API Gateway are positioned to support these evolving needs, offering flexibility, scalability, and ease of integration across diverse environments and architectures.

Kong has implemented U-First Friday, a monthly initiative dedicated to fostering professional development and skill enhancement among its employees, known as Kongers. This program, held on the first Friday of every month, encourages employees to dedicate at least four hours to learning new skills, supported by resources like Udemy, and involves creating individual growth plans with their managers to identify areas for improvement. By blocking time on their calendars, Kongers signal the priority of learning and personal growth, promoting a culture of curiosity and continuous learning. The initiative also encourages social and peer-to-peer learning, leveraging the company's remote-first setup to facilitate knowledge sharing across global teams. Through U-First Friday, Kong aims to empower its employees, aligning with its core value of being an "Explorer," which emphasizes curiosity, boldness, and ongoing improvement.