// blog

DevOps news, explained

One piece of infrastructure news a day: what changed, why it matters to a team without a platform team, and what to do about it this week.

22 Oct 2026 · Hacker News

PHP security release fixes two critical RCE flaws: update now

PHP 8.5.12, 8.4.27, 8.3.36 and 8.2.35 fix two critical remote code execution bugs. What small teams should check and patch this week.

Google Cloud outage hits European banks: check your own exposure

A reported Google Cloud outage disrupted banking and other services in Europe. What small teams should check about single-provider dependence this week.

Distroless images and the gap in shrinking containers

The New Stack looks at the security gap in distroless images and the case for bottom-up OCI containers. What small teams should check this week.

Wiz Runtime Sensor arrives as a managed GKE add-on

Google has put the Wiz Runtime Sensor into GKE as a managed add-on, in preview. What small teams should check before turning on runtime detection.

AI agents that relearn your codebase every session cost money

A New Stack piece on Uvi highlights agents re-reading Slack, Jira and old PRs each time. What that costs a small team and what to check.

Running n8n on EKS: lessons from a Helm chart migration

A CNCF write-up on moving production n8n on EKS to the official Helm chart shows why state, storage and rollback plans matter before any cutover.

BeyondTrust Remote Support zero-day: patch and check

A suspected espionage group exploited a BeyondTrust Remote Support flaw as a zero-day. What small teams should patch, review and rotate this week.

AI agent "hacks" are mostly containment and monitoring gaps

Experts say recent AI agent incidents come from weak sandboxing and little monitoring. What that means if you run agents in your own environment.

Security Hub adds resource criticality to exposure ranking

AWS Security Hub can now weigh exposures by how critical a resource is. What it means for small teams, and what to set up this week.

GPUs on Kubernetes: you build the cost view yourself

The New Stack argues GPUs on Kubernetes are bring-your-own-everything, starting with the bill. What small teams should check before the invoice arrives.

Security Hub can now export findings to S3 as CSV or JSON

AWS Security Hub now exports findings to S3 in CSV or JSON (OCSF). What it means for audit evidence and compliance reporting in small teams.

Kubernetes and cgroup v1: check your nodes before upgrading

The New Stack says Kubernetes on cgroup v1 is finished. Here is how a small team can check its nodes and plan the move to cgroup v2 calmly.

Who owns NIS2 and DORA on a Kubernetes platform?

A CNCF article argues NIS2 and DORA work stalls when it lands on the security board. Put artifacts in sprints instead. Here is what to do.

AI SRE agents: who decides during an incident?

AI incident tools help troubleshoot, but the decision to act is often still human. What small teams should settle before trusting an agent in production.

Blogging anti-patterns, applied to alerts and runbooks

A popular post on software blogging mistakes holds lessons for runbooks and alert text. Here is how small teams can apply them this week.

Kubernetes node swap: fitting more pods on each node

Kubernetes node swap has been GA since v1.34. New benchmarks show higher pod density with NVMe-backed swap. What to check before you try it.

XCOR and AI observability: it still pages engineers

Palo Alto Networks launched XCOR for AI-driven outage tracing, yet humans still get paged. What small teams should do to be ready for incidents.

GKE CPU startup boost: faster pod starts without extra CPU

Google has put CPU startup boost for GKE into preview. Here is what it does, who benefits, and what a small team should check this week.

AI speeds up exploits: why vulnerability spreadsheets fall behind

The New Stack argues AI is speeding up exploits and manual vulnerability tracking can't keep pace. What a small team can change this week.

OpenTofu Day 2026: what a Terraform migration involves

OpenTofu Day returns on 9 November at KubeCon North America. We explain what it means if you run Terraform and are weighing a move to OpenTofu.

GuardDuty Runtime Monitoring moves under Security Hub billing

AWS now bills GuardDuty Runtime Monitoring through the Security Hub Threat Analytics plan. Here is what changes on your bill and what to check this week.

Atlassian's incident detection rebuild: lessons for small teams

Atlassian cut event-to-metric time from 40+ seconds to under 10 on Kafka, Flink and OpenTelemetry. What small teams can take from it, including the silence…

Tell us where it hurts.

The first call is free — and it's with an engineer.

Book a call Send an email