gRPC: 5 Years Later, Is It Still Worth It? (8 minute read)
Torq chose gRPC over OpenAPI/Swagger due to past difficulties with Go, leading to more efficient microservice communication and code consistency. Enhancements by buf.build have further simplified development and dependency management. Torq's use of tools like Linkerd and connectrpc has also addressed initial gRPC-web challenges, optimizing both security and service interaction.
Charlotte Tilbury's move from CloudFormation to Terraform (7 minute read)
Charlotte Tilbury's platform team transitioned from AWS CloudFormation to Terraform for its infrastructure management due to an appreciation of Terraform's drift detection and idempotency capabilities. The migration involved importing existing resources, modularizing code, and leveraging HCP Terraform for a user-friendly CI/CD experience. It ultimately enhanced scalability and security while empowering developers.
Tips to optimize and secure Azure Functions (5 minute read)
Organizations often adopt Azure Functions as part of their cloud modernization strategy for its serverless, event-driven capabilities and cost-efficient pay-per-use model. However, optimizing performance, managing costs, and ensuring security in Azure Functions requires careful selection of hosting plans, efficient design, and proactive monitoring to mitigate potential challenges like DDoS attacks and cold starts.
The Fundamental Law Of Software Dependencies (2 minute read)
This article highlights the importance of incorporating checksums in software source code for all dependencies, including source code, third-party libraries, and compilers, to ensure security and reproducibility. It advocates for using a content-addressed system and lockfiles with checksums to reduce reliance on external distributors and improve dependency management.
GenOps: learning from the world of microservices and traditional DevOps (6 minute read)
Managing generative AI applications presents unique challenges compared to traditional data and AI teams, suggesting the need for a dedicated ‘GenOps' team. Unlike microservices, generative AI involves complex, non-deterministic agents requiring new operational strategies, including centralized model and tool management, and robust security measures.
Leveraging Azure native tooling to hunt Kubernetes security issues (5 minute read)
Container binary drift occurs when a running container deviates from its original image, potentially introducing security risks. This post is part of a three-part series on automating security hunts, leveraging AI, and monitoring Kubernetes security on Azure. It focuses on detecting and preventing binary drift using Microsoft's Defender for Cloud and XDR tools.