TLDR DevOps 2025-11-14
Agent Sandbox π¦, Postgres Internals π, Kubernetes Contexts βοΈ
Introducing Agent Sandbox: Strong guardrails for agentic AI on Kubernetes and GKE (4 minute read)
Google's Agent Sandbox is a new Kubernetes primitive designed for secure, scalable agent code execution with strong isolation built on gVisor and Kata Containers. Integrated with GKE, it improves AI agent performance through features like pre-warmed sandboxes and Pod Snapshots, enabling faster startup times and efficient compute utilization for AI and reinforcement learning workloads.
Helm Marks 10 Years With Release of Version 4 (2 minute read)
Helm 4, released on Helm's 10th anniversary, simplifies and secures Kubernetes application deployment with improved SDKs, a modern plugin system, and support for multi-cluster and CI/CD workflows. It retains familiar interfaces while enabling future chart features, enhancing performance and developer flexibility across cloud-native environments.
Docker Desktop 4.50: Indispensable for Daily Development (2 minute read)
Docker Desktop 4.50 enhances development productivity with free debugging tools, deeper IDE integration, seamless Kubernetes deployment, and enterprise-grade security controls that do not slow workflows. It also simplifies AI-native development through accessible Model Context Protocol integrations, dynamic MCPs, and guided onboarding, enabling teams to build, test, and deploy applications efficiently at scale.
Wrangling Kubernetes contexts (3 minute read)
Kubernetes contexts are dangerously easy to overlook because a single hidden line in ~/.kube/config controls which cluster every kubectl command targets, making accidental production changes far too easy. A safer pattern is to keep only the development config as the default and switch explicitly to production by setting KUBECONFIG via shell aliases, so every risky action must be intentionally prefixed rather than relying on a global context.
Why keep your index set lean (8 minute read)
Extra indexes in Postgres slow down writes (every INSERT/UPDATE must update all of them), increase planner overhead, and compete for cache, hurting read performance. They also waste disk space, add autovacuum work, and generate more WAL. Unused or redundant indexes should be dropped and bloated ones rebuilt to keep databases fast and healthy.
Postgres Internals Hiding in Plain Sight (9 minute read)
Postgres exposes a rich set of internal catalogsβaccessible through psql commands like \d, \di, and \dx, and views such as pg_stat_activity, pg_stat_statements, and pg_locksβthat reveal live activity, performance stats, locks, configuration, and role information. Beneath these views are core catalog tables (pg_class, pg_attribute, pg_type, pg_proc, and pg_stats) that store the database's own metadata. Using psql -E or ECHO_HIDDEN lets you see the exact SQL behind psql commands to explore these internals more deeply.
π¨βπ»
Resources & Tools
One platform. Scalable AI agents. Real results. See why OutSystems is the #1 AI-powered low-code platform (Sponsor)
OutSystems Agent Workbench simplifies agentic AI development for companies around the world with the speed and flexibility of a low-code approach. Trusted by Axos Bank, Thermo Fisher Scientific, and The Arch Company, it provides a single, unified platform for scaling complex agentic systems. Deploy agents across dev, test, and production with one-click publishing.
Schedule your demo Strix (GitHub Repo)
Strix is open-source AI agent that emulates hackers by dynamically running code to identify and validate vulnerabilities, providing fast security testing for developers. GPT-5 and Claude Sonnet 4.5 are recommended for optimal use. Strix can be integrated into CI/CD pipelines.
Ingress NGINX Retirement: What You Need to Know (3 minute read)
Due to maintenance challenges and security concerns, Kubernetes SIG Network and the Security Response Committee are retiring Ingress NGINX, with best-effort maintenance until March 2026. Users are advised to migrate to alternatives like Gateway API, as no further bug fixes or security updates will be provided after the retirement date, though existing deployments will continue to function.
Redis Critical Remote Code Execution Vulnerability Discovered after 13 Years (3 minute read)
CVE-2025-49844 is a critical use-after-free vulnerability in Redis' Lua scripting engine that allows authenticated attackers to escape the Lua sandbox and execute arbitrary code on affected Redis and Valkey instances. The flaw, originating from a 13-year-old memory corruption bug, was patched in recent releases.
Homebrew 5.0.0 (3 minute read)
Homebrew 5.0.0 introduces default parallel downloads, official Linux ARM64 support, and a multi-year deprecation path for Intel macOS and non-codesigned casks. It also adds numerous improvements to brew commands, tighter macOS security alignment, and a new internal JSON API (opt-in). The update consolidates all external commands into Homebrew's primary repositories.
Get our free daily newsletter with curated tools π», trends π, and insights π‘, for DevOps Engineers π¨βπ»
Join 340,000 readers for
one daily email