December 2022 Summaries

AWS Lambda External Extensions are a feature that allows developers to add code to their Lambda handler functions in separate processes, improving maintainability and performance. This separation of concerns enables developers to implement features like logging without modifying the original Lambda function code, reducing maintenance work and increasing resilience. By running external extensions in parallel with the main Lambda function, these extensions can leverage multiple CPUs, leading to improved performance and reduced impact on user experience. Additionally, external extensions can execute code after the main function returns, allowing for additional processing without affecting the user. Overall, AWS Lambda External Extensions are a powerful tool for developers to improve the scalability, reliability, and maintainability of their Lambda functions.

Building an observability solution tailored to a serverless system requires data, and a lot of it. Logs, metrics, and traces are required to ensure the system can answer as many questions as possible about how it’s working, whether you’re repairing or optimizing. Each source contributes a different perspective on how your application is performing and, when analyzed together, give a complete picture of its health. Choosing an observability solution designed specifically for serverless contexts will help you reap all the benefits that serverless offers, while avoiding generic monitoring tools that might drown actionable information in a sea of metrics. A centralized dashboard provides a single pane of glass experience to monitor your entire stack from serverless functions to APIs and other services in one place, without the need to switch context or tools. Building an observability solution can be a major hidden cost in running a serverless application, but with the right tool, such as Lumigo, you can minimize these costs by focusing on service-specific data, implementing automatic instrumentation, using distributed tracing, and configuring log retention correctly.

AWS re:Invent was held in Las Vegas with approximately 50,000 attendees and featured various announcements spanning new features and improvements to AWS services. The new feature of AWS Lambda SnapStart reduces cold start delays for Java-based functions by creating an immutable, encrypted snapshot of the container after initialization, allowing future invocations to launch from a warm start. Additionally, AWS Application Composer visualizes serverless architectures, while Amazon CloudWatch Logs data protection helps keep customers and certifications safe by masking sensitive information in logs. Cross-Account Observability tracks the CloudWatch logs of multiple accounts in one place, and Amazon Inspector expands its vulnerability scans to include AWS Lambda functions and layers. Other notable announcements include Amazon OpenSearch Serverless, Amazon SageMaker Serverless Inference, and serverless data and analytics options using EMR, Kinesis, MSK, and Redshift.

Smoke testing in production refers to the practice of running tests on actual data from real users in a production environment, extending other testing methods like unit and integration testing. This approach allows for quick assessment of an application's status by running end-to-end tests targeting significant user flows, which are collections of basic and main functionality in the system. Smoke testing is typically run after new deployments and at regular intervals, consisting of many small and quick-to-execute testing blocks to identify production issues as fast as possible. Implementing smoke testing in production involves automating critical flows using tools like Cypress, creating a CI platform such as CircleCI, scheduling pipelines, and monitoring results with services like AWS CloudWatch. By following best practices, such as avoiding data pollution, flakiness, and false positive alarms, developers can ensure that their application behaves as expected in the production environment.