TLDR DevOps 2025-06-16

Cloudflare experienced a two-and-a-half-hour global outage on June 12 due to a failure in the third-party cloud-backed storage infrastructure that supports its Workers KV service. Critical services including WARP, Access, Gateway, Workers AI, and parts of the Dashboard were disrupted, though no data was lost and core network services remained unaffected.

AWS is launching the European Sovereign Cloud with a fully EU-based governance structure, independent infrastructure, and operations entirely managed within the EU to meet strict digital sovereignty requirements. The cloud will offer a full suite of services, including AI and security tools. It is supported by a €7.8 billion investment and designed to maintain performance while ensuring compliance with European laws and standards.

GitLab reduced repository backup times from 48 hours to 41 minutes by identifying and fixing a performance bottleneck in a 15-year-old Git function. By replacing nested loops with a map-based deduplication method, it improved scalability, lowered resource usage, and enabled frequent backups for large repositories.

Amazon Q Developer's IDE plugins for Visual Studio Code and JetBrains now support Model Context Protocol (MCP), allowing developers to connect external tools and MCP servers for more context-aware responses.

Distributed tracing helps increase developer velocity and deliver meaningful innovation. Read this technical blog by Dynatrace to see how you can use distributed tracing to visualize and understand your system, analyze and fix problems fast, contextualize telemetry data, and build automation workflows.

Discover how Dynatrace delivers a frictionless observability experience for developers. 👉Learn more

Sourcerer is a CLI-based cloud storage explorer that gives developers and DevOps engineers a unified interface to manage files across multiple cloud providers like GCP Storage, Azure Storage, AWS S3, and S3-compatible services. The tool allows for the management of cloud storage directly from the terminal.

Kyverno and OPA Gatekeeper are policy enforcement tools that secure Kubernetes clusters by preventing misconfigurations like the use of unversioned image tags. Kyverno offers a simpler, YAML-based approach tailored for Kubernetes users, while OPA Gatekeeper provides greater flexibility through the Rego language for complex, cross-platform policy needs.

Netflix has built UDA (Unified Data Architecture) as a foundation for connected data in Content Engineering, which allows teams to model domains once and represent them consistently across systems to power automation, discoverability, and semantic interoperability.

A global outage affected numerous Google Cloud and Workspace products on June 12. The outage was caused by a buggy quota policy update that triggered a crash loop in a critical API management service, Service Control. The issue was resolved within three hours for most regions, but us-central1 experienced delayed recovery due to infrastructure overload. Google has committed to major architectural and communication improvements to prevent recurrence.

Amazon ECS now allows users to update capacity providers for existing services, enabling changes to the compute configuration without service disruption or needing to recreate services.

The first OpenFeature-native platform, built on edge architecture for fast, scalable rollouts. Helps teams ship faster, reduce risk, and experiment with confidence. Start free.

DigitalOcean customers now have access to AMD Instinct MI300X GPUs as DigitalOcean GPU Droplets.

Grafana Tempo 2.8 was released with new TraceQL features, performance enhancements like compactor pooling, and some breaking changes including the removal of deprecated serverless features.

New Relic now supports the Model Context Protocol (MCP) within its AI Monitoring solution, integrated with its Application Performance Monitoring (APM), to give insights into the lifecycle of an MCP request.