Search the archive

CUTTING PROVIDER MEMORY BY 80%: CROSSPLANE-RUNTIME V2.4 AND CLIENT CACHING

Crossplane providers running crossplane-runtime v2.4.0 see memory usage drop by roughly 80-95% on clusters with around 2,000 CRDs.

whitepaper_

WHAT'S NEW IN AI INFRASTRUCTURE AND ORCHESTRATION IN SEPTEMBER

Google Cloud's September roundup highlights GKE Agent Substrate with 10x sandbox density, Pod snapshots cutting inference startup by 89%, a multi-cluster Inference Gateway, and new Filestore and VM storage options. It also cites analyst rankings in Forrester, Gartner, and SemiAnalysis ClusterMax.

EVERYTHING WE LAUNCHED DURING BIRTHDAY WEEK 2026

Cloudflare shipped 46 announcements during its Birthday Week 2026, including a new command-line interface called cf and an open-source pipeline named Forge. The company intends to become a public certificate authority and is preparing its cryptography for the post-quantum era. Cloudflare also launched a beta Monetization Gateway that lets creators charge AI agents for content using HTTP 402.

PROACTIVE DEFENSE: HARDENING CODE PIPELINES AND CI/CD INFRASTRUCTURE

Ephemeral runners, short-lived OIDC credentials, immutable dependency references, and provenance verification limit opportunities for software supply-chain compromise. Isolating build caches by branch privilege, blocking untrusted pull requests from release secrets, and quarantining incoming dependencies help prevent malicious code from reaching trusted builds.

WHEN AI AGENTS SWARM, CAN BANKS KEEP UP?

Banks face new risks as AI agents act autonomously at machine speed, and individual controls alone may miss unexpected behavior. Elastic argues for unified data across logs, metrics, traces, threat intelligence, audit trails, and security operations that combine human judgment with automated investigation.

SCALING KUBERNETES WORKLOADS WITH NODE SWAP

Kubernetes support for running nodes with swap enabled reached General Availability in v1.34. Tests using Local SSDs for swap showed density gains of up to 3× across workloads like kernel builds and isolated Python sandboxes. One benchmark scaled concurrent Python sessions from 80 to 240 by offloading dormant memory to disk.

ungated whitepaper_

REA

REA lets users ask an AI agent to investigate app features without source code, explaining how they work down to the native binary level. It includes tools for inspecting JavaScript and Electron apps, .NET assemblies, websites, and native binaries through Hopper, Ghidra, or IDA Pro. Analysis runs locally, and the tool supports agents like Claude Code, Cursor, and GitHub Copilot CLI.

E2E

e2e is an end-to-end testing framework for web and mobile apps. Describe a goal in natural language and an agent drives the app to reach it. Check the result with locators and assertions in the same test.

THE SHIFT TO CGROUP V2 IN KUBERNETES: WHAT YOU NEED TO KNOW

Kubernetes v1.35 defaults to refusing kubelet startup on cgroup v1 nodes, making migration to cgroup v2 an operational upgrade concern. Teams should check kernel and runtime compatibility, align kubelet and runtime cgroup drivers, and update monitoring tools that read cgroup files directly.

HOW FAST IS PYTHON 3.15?

Informal CPU-bound benchmarks of Python 3.15.0rc3 show standard-interpreter performance largely unchanged from 3.14, while the experimental JIT runs roughly 20–28% faster than the standard 3.15 interpreter. Free-threaded builds deliver substantial gains in four-thread tests, but these results cover only Fibonacci and bubble-sort workloads rather than general application performance.

CURSOR USES S3 WAL TO SCALE GIT STORAGE TO MORE THAN 300 PUSHES PER SECOND

Cursor's Continuity treats S3-backed write-ahead logs as the source of truth for Git storage.

"THINK OF IT AS KUBERNETES FOR AGENTS": OPENCLAW LANDS IN THE ENTERPRISE WITH OPENAI, NVIDIA AND RED HAT ON BOARD

OpenClaw Enterprise is an open-source control plane for managing AI agents on Kubernetes launched with backing from OpenAI, Nvidia, and Red Hat.

PULUMI GOOGLE CLOUD PROVIDER VERSION 10.0.0

Pulumi has released version 10 of its Google Cloud provider.

Free. Register for the livestream

LAST WEEK IN KUBERNETES DEVELOPMENT: WEEK ENDING SEPTEMBER 27, 2026

Two medium-severity Kubernetes CVEs were disclosed and patched, SIG Cluster Lifecycle is getting a new chair, and v1.38 entered Enhancements Freeze with 89 of 100 enhancements remaining. Device Binding Conditions graduated to GA, and steering elections close on October 2.

CD FOUNDATION PROJECT UPDATES SEPTEMBER 2026

The CD Foundation's September update covers CDEvents spec v0.5.1, Jenkins finishing five Google Summer of Code projects, and Spinnaker 2026.3.0, which removes AWS SDK1 and Halyard. Ortelius, Screwdriver, and JayeX also reported progress.

INTRODUCING CLOUDFLARE TRACES: FOLLOW REQUESTS THROUGH OUR ENTIRE PLATFORM

Cloudflare Traces is now available in open beta, extending automatic tracing beyond Workers to the rest of the request path. Users can view security rules, transformations, cache decisions, and origin handling in one trace without special configuration or plugins. The feature supports OpenTelemetry and enables users to control sampling rates or trace 100% of specific traffic using Trace Rules.

CLOSED-LOOP INCIDENT RESPONSE: CONNECT AWS DEVOPS AGENT TO OPENSEARCH

A new AWS guide explains how to connect AWS DevOps Agent to Amazon OpenSearch Service using the Model Context Protocol (MCP). This setup allows an AI agent to automatically investigate alerts by querying logs and traces, rather than requiring manual intervention from an engineer. Organizations can host the MCP server on Amazon ECS, Amazon Bedrock AgentCore, or use the built-in endpoint on OpenSearch version 3.3+.

HALVING THE TIME: HOW UBER EATS REBUILT ITS SEARCH PIPELINE

Uber cut end-to-end Eats search latency in half by removing low-value retrieval paths, separating ranking features from presentation data, and eliminating false dependencies between pipeline stages. Additional gains came from pagination, concurrent rendering, request hedging, compact embeddings, and fixing service-mesh connection bottlenecks.

scaled testing across hundreds of engineers and agents.

IMPECCABLE

Impeccable is a design skill for AI coding agents that includes 24 commands and 61 deterministic detector rules for identifying common problems in AI-generated frontend design. The tool works with editors including Cursor, Claude Code, VS Code, and Grok Build.

T3CODE

T3 Code is an "agent harness control surface" that manages AI agents from Claude, Codex, Cursor, and other platforms. Users can control these agents through mobile, web, and desktop apps.

REBALANCER

Meta open-sourced Rebalancer, a library for assigning tasks, shards, servers, and traffic while balancing resource limits, placement constraints, and optimization objectives. It supports local-search and mixed-integer solvers, plus a Dockerized Explorer interface for debugging why assignments were made and which constraints are binding.

BEYOND OBSERVABILITY: EVOLVING PRODUCTION OPERATIONS IN THE AGE OF AI

Panelists from Genesys, Netflix, and Groundcover discuss using AI agents for incident response and observability, stressing human accountability, machine-readable high-cardinality telemetry, sandboxed rollouts, and organizational context. They advise starting small in greenfield environments and documenting jobs to be done for agents.

INTRODUCING WEB SEARCH API VIA AI GATEWAY

Cloudflare partnered with web search providers Ceramic.ai, Exa, and Linkup to add web search capabilities to its AI Gateway. The feature lets agents retrieve live data instead of guessing URLs. Web search partners must comply with Cloudflare's "Verified bots" requirements and respect robots.txt.

WHAT IS IT AUTOMATION? A COMPLETE GUIDE

IT automation replaces manual data center and cloud work with repeatable scripted processes.

HIGH AVAILABILITY IS NOT RESILIENCE: WHY CLOUD SYSTEMS FAIL WHEN IT MATTERS MOST

A TLS 1.3 upgrade broke Route 53 health checks and silently rerouted a region's traffic, showing that high availability differs from resilience.

Own your AI

BUILDING A POST-QUANTUM CERTIFICATE AUTHORITY WITH MERKLE TREE CERTIFICATES

Cloudflare is launching a certificate authority that will support Merkle Tree Certificates (MTCs), targeting early 2027 for inclusion in Chrome's Quantum-resistant Root Store. The company will provide standard MTC issuance at no cost. MTCs batch certificates into an append-only Merkle tree to address the scaling issues of post-quantum cryptography, which can increase log storage requirements by 40 times. Cloudflare successfully tested the technology with Chrome, serving billions of MTCs and finding that median performance was 9% faster than classical signature chains.

AMAZON S3 VECTORS NOW SUPPORTS METADATA PRE-FILTERING FOR HIGHER RECALL ON FILTERED SEARCHES

AWS has introduced metadata pre-filtering for Amazon S3 Vectors to improve recall on filtered queries. The feature evaluates metadata filters before running a similarity search and supports up to 100 filter constraints per query. On highly selective filters, pre-filtering returns up to five times more matching vectors than previous methods. It is available immediately at no additional cost and works on existing indexes without requiring data re-ingestion.

CLOUDFLARE CONTAINERS, REBUILT TO SCALE AGENT SANDBOXES

Cloudflare has rearchitected its Containers service to handle agent workloads, allowing code to select the image and instance type at runtime. The update introduces a new scheduling policy that starts containers six times faster, with median startup times falling from over four seconds to 648 milliseconds in independent benchmarks. A filesystem snapshots feature is also available in public beta to let workspaces save and restore their state.

HOW TO JUSTIFY INVESTMENT IN A PLATFORM

Gartner forecasts that 40% of companies will downgrade autonomous AI agents by 2027 due to governance gaps found after production incidents. Google's DORA 2025 research found that 90% of organizations already use an internal platform, and 76% have dedicated teams for them. Broadcom research shows 97% of IT leaders believe some public cloud spending is wasted, with cost now exceeding security as the primary concern.

WE TESTED OUR OWN WAF WITH FRONTIER AI MODELS. HERE'S WHAT WE FOUND

Cloudflare built a dynamic testing tool that uses large language models to iterate attack payloads against its Web Application Firewall. The system ran 1,107 attempts across six attack categories and recorded 49 findings that required human review. Cloudflare used these results to release new detections for SSRF attacks and improve an existing cloud metadata rule.

the Fortune 50 trust RunMyJobs

DBX

DBX is a 25 MB cross-platform database client that supports over 100 databases, including MySQL, PostgreSQL, SQLite, Redis, and MongoDB. It runs as a native desktop app on macOS, Windows, and Linux, or via Docker and web deployments, without requiring Java, Python, or bundled Chromium. The tool includes an AI assistant that generates SQL from natural language and an MCP server that lets coding agents query databases.

HYPERFRAMES

HyperFrames is an open-source framework that converts HTML, CSS, and seekable animations into deterministic MP4 videos. The toolkit ships 21 skills that teach AI coding agents the full production loop, from planning and writing valid HTML to rendering with FFmpeg. It renders video by seeking frames in headless Chrome to ensure the same input always produces the same output.

CUT AI AGENT COST AND IMPROVE ACCURACY WITH CODE EXECUTION IN THE DATADOG MCP SERVER

Datadog's new Code Execution feature lets AI agents run sandboxed JavaScript against its MCP Server, querying, joining, and filtering data across multiple APIs before returning only relevant evidence to the model. Testing across four models showed accuracy gains up to 21.8 points alongside 73% fewer input tokens and 40% fewer tool calls.

THE 4 LEVELS OF AGENTIC SOFTWARE DEVELOPMENT

Enterprise AI productivity gains depend on platform maturity rather than model choice, with four levels running from humans approving every change to fully autonomous agents. Most organizations sit between levels one and two, and shifting validation from a single human gate to a continuous loop marks the hardest, highest payoff transition.

it live for free on October 6