Tag
Google发布了《Site Reliability Engineering》第二版，距2016年首版十年之际，全面更新了SLO、可观测性、自动化、容量管理等六大实践领域，并深入探讨了AI与SRE的双向关系——AI既是运维对象也是运维工具。
Uber engineers detail the design of their MCP Gateway, a centralized microservice platform that unifies how hundreds of teams expose existing HTTP, gRPC, and TChannel services as MCP tools for AI agents — now hosting 800+ MCP servers and 5000+ tools internally.
DronaHQ MCP is introduced as a bridge between AI coding assistants and backend systems, providing scoped access without exposing raw credentials. It handles permissions, secrets, auth, audit trails, and deployment so AI agents can safely build application logic in production.
An analysis of how platform engineering must evolve for AI agents, shifting from rigid golden paths to composable, API-first building blocks that support non-human identities, scoped permissions, and audit trails.
An opinion piece arguing that platform engineering and code reuse remain valuable even with AI coding tools, because tokens cost money and reuse provides leverage. Cites Martin Fowler's similar argument about refactoring.
Tigera launches Lynx, a Kubernetes-native unified control plane for managing AI agents at scale, providing identity, policy enforcement, and real-time visibility across multi-cluster environments.
The article explores how platform engineering teams must adapt to AI workloads by managing costs and risks without overhauling existing infrastructure, highlighting the shift from traditional DevOps to a new paradigm.
This article explores the true meaning of Forward Deployed Engineering (FDE) in AI deployment, emphasizing that FDE is not simply about API calls or building agents, but rather a systematic engineering approach geared toward production deployment, including business translation, system design, platform integration, production operations, and capability accumulation.
A podcast conversation with CI/CD expert Robert Erez explores how AI is reshaping CI/CD practices, the nuances of GitOps, and trends like ephemeral environments and progressive delivery.
OpenRun introduces a service binding feature that automates database access for applications, creating isolated database environments without manual configuration.
The article highlights a feature of Google Cloud Run that simplifies building zonal redundancy, noting that few other platforms offer this capability.