Stanislav DeviatovArticles
The article reviews various market data solutions for futures trading, including Barchart, TickAPI, Interactive Brokers, TradeStation, YCharts, Databento, Polygon.io, thetadata, Pitchbook, Binance, and AlphaSense. Each platform offers unique features, from comprehensive data sets and deep historical data to flexible pricing and AI-driven insights. The best solution will depend on the specific needs and use case of the user.
π 5 REST APIs With Great Developer Experience
The article highlights the importance of Developer Experience (DX) in determining the adoption and usability of an API. It cites examples of five APIs with high-quality DX: Stripe, known for its comprehensive API documentation; Slack, which offers an array of SDKs for various programming languages; Pinterest, which provides multiple versions of its API; PayPal, recognized for its well-organized API documentation and developer community; and Sabre, praised for its extensive sample code. The article emphasizes that good DX encompasses easy onboarding, thorough documentation, and a supportive developer community.
π AsyncSLA: Towards a Service Level Agreement for Asynchronous Services
It introduces AsyncSLA, a domain-specific language for specifying service level agreements for asynchronous services, and a quality model for asynchronous services. The authors have also developed a tool that integrates AsyncSLA with the AsyncAPI specification, aiding in the definition of service level agreements for asynchronous services.
π Confluent HTTP/REST proxy vs. Gravitee Kafka proxy
The article compares different options for proxying Kafka, focusing on Confluent HTTP/REST proxy, Confluent REST API, and Gravitee Kafka Support. It discusses how these solutions address common challenges faced by organizations implementing Kafka, such as integrating with existing applications and providing secure access to real-time data. The article highlights the strengths of each solution, including Confluent's extensive Kafka REST APIs and Gravitee's flexible architecture and support for multiple event broker systems. It concludes by emphasizing Gravitee's mission to unify synchronous and asynchronous data across any gateway and broker technology.
π Data integration patterns explained
The article discusses the importance of data integration patterns in managing data from multiple sources. It explains key activities in data integration like extraction, transformation, and loading, and outlines various techniques such as Extract, Transform, Load (ETL) and Extract, Load, Transform (ELT). The article also presents different data integration patterns including migration, broadcast, bi-directional sync, correlation, and aggregation. It emphasizes that these patterns can enhance operational efficiency, improve data quality, and provide better customer insights.
The article discusses TypeSpec, an API description language created by Microsoft for developers, which aims to differentiate itself from similar description languages. TypeSpec was designed to meet the needs of a large and varied developer community and includes several features to help them embrace API-first. The article highlights that TypeSpec acts as a layer below API description languages like OpenAPI. It also mentions that TypeSpec provides a lightweight language for describing and designing APIs, and it's easy to integrate into an OpenAPI toolchain. The article further explains that TypeSpec allows API guidelines to be expressed in reusable code. The article suggests that while TypeSpec has its advantages, its suitability for a particular user would depend on their specific needs and preferences.
π Fixing duplicate API requests
The article discusses the issue of duplicate API requests in distributed systems and introduces the concept of idempotency as a solution. Idempotency refers to operations that can be applied multiple times without changing the result beyond the initial application. The author refers to the IETF specification "The Idempotency-Key HTTP Header Field", which suggests using a unique key sent by the client along with the request. If the server already has a record of the key-request pair, it discards the request; if not, it stores the pair. The article also outlines potential error scenarios and the corresponding HTTP status codes. The author concludes by suggesting that the behavior described could be factored into an API Gateway, and plans to explore implementing this behavior in Apache APISIX in a future post.
π How To Create An API Business Product
The article reviews the book "Building an API Product" by Bruno Pedro, which offers a comprehensive guide to treating APIs as business products. The book covers the importance of conducting market research, delivering exceptional user experience, understanding business value, and familiarizing oneself with common monetization models. It also emphasizes the importance of spending time on API design, developing an API strategy, and calculating expenses at different stages of the API lifecycle. The article highlights that APIs are becoming increasingly important in today's tech landscape, not just for developers but also for other roles within an organization, and can open up businesses to a global marketplace.
π Navigating backwards compatibility in GraphQL
The article discusses the importance of maintaining backwards compatibility in GraphQL, especially when two systems need to communicate seamlessly as their shared interface evolves. The author shares his experiences and challenges faced while managing GraphQL at Jam, a Chrome extension for bug reporting. The author emphasizes the need for a robust approach to maintaining backwards compatibility due to potential disruptions caused by changes in the service. The article further explores various strategies for maintaining backwards compatibility, including versioning GraphQL objects, parameterizing versions, and adopting a versionless API approach. The author also discusses the use of conventions and automation tools like GraphQL Inspector to prevent potential issues. The article concludes by highlighting the importance of balancing innovation and stability in GraphQL to ensure a smooth user experience.
π You probably don't need GraphQL
The author argues that while GraphQL is a powerful tool, it may not be necessary for all use cases. GraphQL was designed as a central intermediary layer between various end-user-facing clients and data sources. It is particularly useful for maintaining backwards compatibility, preventing mobile apps from breaking after API changes, improving loading times by reducing request waterfalls and overfetching, simplifying maintenance and endpoint discovery, and enhancing security and performance. However, for simpler or smaller-scale projects, the author suggests that other solutions may be more appropriate. The author concludes that while GraphQL may not be necessary for all projects, it can be a valuable tool when specific problems arise.
π What is API-as-a-Service?
The article discusses the concept of API-as-a-Service (APIaaS), a technology solution that provides systems and services to develop, deploy, and manage APIs at scale. The author explains that APIs, as value-adding systems, enable businesses to connect and translate disparate systems, creating a unified ecosystem. However, developing APIs can be costly and complex. APIaaS offers a solution by providing functions and resources as a service, reducing costs and simplifying the process. The author discusses the benefits of APIaaS, such as improved efficiency, reduced time-to-market, and better utilization of resources. However, he also notes potential drawbacks, including loss of control, security concerns, and cost variability. The author concludes that the decision to implement APIaaS should be based on the organization's need for efficiency versus control.
Apache APISIX
π Implementing the Idempotency-Key specification on Apache APISIX
The article discusses the implementation of the IETF Idempotency-Key specification using Apache APISIX. The Idempotency-Key specification aims to avoid duplicated requests by allowing the client to send a unique key with the request. If the server doesn't know the key, it proceeds as usual and then stores the response. If the server knows the key, it immediately returns the stored response. The article provides a step-by-step guide on how to implement this specification with Apache APISIX, including the creation of a plugin, configuring the Redis client, implementing the nominal path, and managing error paths. The author concludes by stating that while the current implementation is simple, it can be improved by adding automated tests, configuring Redis on a per-route basis, and using other key/value stores.
π Five ways to pass parameters to Apache APISIX
The article discusses how to pass parameters to Apache APISIX, an open-source API gateway. It covers five methods: path variables, query parameters, request headers, request body, and cookies. For each method, the author provides the corresponding API or method in APISIX or Nginx for accessing the parameters. The author concludes by summarizing these methods and acknowledging a reviewer's contributions.
Apache Camel
π Camel integration quarterly digest: Q1 2024
Red Hat has launched a quarterly newsletter focusing on Apache Camel, an open-source integration framework. The first edition announces the Red Hat build of Apache Camel 4.0, featuring the Kaoto visual designer and the Hawtio diagnostic console. The newsletter also shares articles, demos, and upcoming events related to Apache Camel. The next major release, the Red Hat build of Apache Camel 4.4, is expected in May.
Apache Kafka
π Schema Registry Clients in Action
The article provides an in-depth explanation of how the Schema Registry works in conjunction with Apache Kafka, focusing on the behind-the-scenes processes that occur when messages are serialized and deserialized. The Schema Registry is a component that manages and stores schemas for data exchanged between Kafka producers and consumers, ensuring data quality, compatibility, and consistency. The article also discusses the use of data quality rules on the schema, which are crucial for maintaining high data integrity and simplifying schema evolution. Lastly, the article provides a demonstration of how to use the Confluent Schema Registry without Apache Kafka, emphasizing that a producer or consumer client independently communicates with two separate systems: Apache Kafka and the Schema Registry.
Gravitee
π Gravitee 4.3 platform update: enhanced documentation support, new TCP proxy UI, and more
The article discusses the updates in the Gravitee 4.3 platform which aims to enhance API management. The platform introduces additional support for asynchronous and event APIs, new GraphQL support, TCP proxy enhancements, and more. Key features include major API documentation enhancements, improvements to v4 APIs, updated entrypoint and endpoint configuration, new policies and policy enhancements, and full TCP Proxy support. The platform also introduces new multi-factor authentication (MFA) capabilities, audit log improvements, and stricter management of signing certificates. These updates aim to improve functionality, security, and user experience in API management.
π Securing and Exposing gRPC Services with Gravitee: A Hands-On Guide
The article provides a guide on using Gravitee 4.3 for managing gRPC services. It explains that gRPC is a high-performance, open-source Remote Procedure Call (RPC) framework developed by Google. The author provides instructions on creating a gRPC proxy API in Gravitee API Manager, testing the proxy, securing gRPC calls with an API Key, and accessing gRPC service analytics. The article emphasizes that Gravitee allows for efficient API management, enhancing their security and performance.
Mulesoft
π Strategy for AnyPoint Platform implementation across multiple businesses
The article discusses strategies for implementing MuleSoft's AnyPoint Platform across multiple businesses within a large organization. The author highlights the importance of maintaining a balance between business independence and central control. He discusses various implementation options, including separate implementations for each business, single implementations for business sectors, and central management of the platform across the entire organization. The author suggests that the choice of implementation should depend on factors such as security requirements, the degree of separation between business sectors, and the specific needs of each business. He concludes by emphasizing the importance of making decisions based on experience and the principle that 'perfection is the enemy of progress'.
Acquisitions
π€ Akamai Acquires Noname Security: A Boon for API Security?
The API security landscape witnessed a significant shift recently with the announcement of the acquisition of Noname Security by content delivery network (CDN) giant Akamai Technologies.
π€ Announcing Postman has acquired Orbit
Postman has announced its acquisition of Orbit, a tool for growing developer communities. The integration aims to enhance collaboration on the Postman platform, allowing API distributors to expand their communities, increase API usage, and receive direct user feedback.
Releases
π Apache Camel K 2.3.0
Apache Camel community has released Camel K 2.3.0, which introduces several stability enhancements, a new default runtime, multi-architecture manifest support, Strimzi Kafka support, and the new IntegrationProfile custom resource. The update also includes a new publishing strategy, builder annotation support, and the deprecation of some features.
π Camunda 8.5
Camunda has released a new component, Camunda Console Self-Managed, designed to enhance the management and monitoring of Camunda installations within enterprise environments. This release also includes simplified installation and configuration, support for Amazon OpenSearch, enhanced applications configurability, improved documentation structure, and support for Zeebe Rolling Updates. The update also introduces a new Db2 for iSeries connector, a shift to Java 17 as a compile-time baseline, and the introduction of the Asynchronous Embedded Engine model. Other improvements include enhanced availability and operations, new features and capabilities in the Web Modeler, and updates to the Tasklist User Interface.
π Debezium 2.6.0
Notable changes include seven breaking changes, the introduction of a new connector for IBM's Db2 iSeries/AS400, a shift to Java 17 as a compile-time baseline, and the introduction of the Asynchronous Embedded Engine model. The release also includes enhancements to the existing connectors, new payload formats, and improved event timestamp precision.
Books
π Postman Cookbook: Hand-picked Solutions and Techniques across API Design, Testing, Performance, Networking, Kubernetes and Integration by Oliver James
The book covers a wide range of topics, from the basics of web-based communication to advanced subjects like custom protocol handling, network troubleshooting, and proxy management. It also discusses the influence of security measures, performance optimization, and scaling solutions. By the end of the book, readers will have a thorough understanding of API development with Postman and will be able to handle modern computing challenges confidently. The book also includes sections on real-time API connectivity, protecting API endpoints, network optimization, API testing in Kubernetes, and advanced authentication flows.
