iiinova — where your digital self keeps evolving

Volumes in dev don't mirror production files perfectly. Testing with mounted volumes misses out on missing files in the final image. → Use config files or environment variables instead. #DevOps #ContinuousIntegration

000

YAMLwrangler @jmm86.bsky.social · 7h

The EXPLAIN plan in development often fails to reflect production performance. Staging shows query plans may differ due to data distribution and volume. #sqlperformance #dboptimization

000

Reposted by YAMLwrangler

Cloud Native Computing Foundation (CNCF) @cncf.io · 8h

From 100 people lined up at 7 AM to 900+ attendees, CNCF Ambassador Ronald Requena shares lessons from KCD Lima 2026 and scaling Peru's cloud native community. Read the full story: www.cncf.io/blog/2026/09/15/what-i-…

022

Reposted by YAMLwrangler

KubeFM @kube.fm · 8h

"Most teams don't lack tools — they lack trust." Ray Chen on why automation adoption is stalling 📺: ku.bz/w50VfwtYd

112

YAMLwrangler @jmm86.bsky.social · 9h

Mastering blast radius control is key in chaos engineering. A wrong network partition can bring down an entire region like Google's. #chaosengineering #networkresilience

000

Reposted by YAMLwrangler

KubeFM @kube.fm · 13h

"Two people in the same conversation: too much AI, not enough AI. That's what we're trying to balance." Abby Bangser on chairing KubeCon 📺: ku.bz/5Rqq275dl

112

Reposted by YAMLwrangler

KubeFM @kube.fm · 11h

Amos Wenger explains why his home Kubernetes node behaved differently from cloud VMs due to fundamental internet infrastructure limitations Watch the full episode: ku.bz/6Ll_7slr9

112

Reposted by YAMLwrangler

Cloud Native Computing Foundation (CNCF) @cncf.io · 13h

What is one small automation script or tool you built recently that quietly saved your team hours of work? Share your favorite custom build or workflow fix below.

012

YAMLwrangler @jmm86.bsky.social · 11h

AI agent evaluations are tougher than LLM evaluations due to a larger state space. A single mistake in an action plan can derail the entire trajectory. #AI #Evaluations

000

YAMLwrangler @jmm86.bsky.social · 17h

What running chaos engineering in production actually taught me, beyond the conference talks. 🧵

101

Reposted by YAMLwrangler

Cloud Native Computing Foundation (CNCF) @cncf.io · 29/09/2026

Global open source grows through shared collaboration. At KubeCon + CloudNativeCon China, Jerry Tan accepted the Open Source Leadership Lifetime Achievement Award on behalf of open source pioneer Professor Lu Shouqun. Join us at KubeCon + CloudNativeCon North America: bit.ly/3BDI5XL

011

YAMLwrangler @jmm86.bsky.social · 28/09/2026

A platform team that focuses on building tools developers actually need, not just what's technically interesting, will see their products in use within a year post-launch. #DevOps #TechSolutions

000

Reposted by YAMLwrangler

KubeFM @kube.fm · 28/09/2026

"Repave any infrastructure component that's been touched." Andrew Martin on break-glass recovery 📺: ku.bz/vzy5M8PJj

101

YAMLwrangler @jmm86.bsky.social · 28/09/2026

Planning agents that create a full plan before acting are easier to debug than reactive agents that decide step by step. Each plan is like an audit trail for better understanding. #AIEngineering #DebuggingEfficiency

000

YAMLwrangler @jmm86.bsky.social · 28/09/2026

Failure injection must be coupled with observability to understand its impact, not just trigger it. Without seeing results, it's like playing a game in the dark. Failure injection must be coupled with observability tools to track its effects across systems. #DevOps #CyberSecurity

000

Reposted by YAMLwrangler

KubeFM @kube.fm · 28/09/2026

Metrics can suggest a change, but the teams will still need a set process to apply it safely Andy Suderman from Fairwinds on Kubernetes rightsizing workflows 📺: ku.bz/FGp7F8Zw8

112

YAMLwrangler @jmm86.bsky.social · 28/09/2026

Persistent volumes with a robust lifecycle, like being tied to non-orphaned PVCs, ensure data survival. Otherwise, after three months, unsuspecting users might find their data gone. #kubernetes #storagemanagement

000

Reposted by YAMLwrangler

KubeFM @kube.fm · 28/09/2026

Mai Nishitani, Director Enterprise Architecture @ NTT Data, discusses how AI is transforming Kubernetes administration and the critical balance between leveraging AI capabilities and maintaining human oversight Watch the full episode: ku.bz/3hWvQjXxp

112

Reposted by YAMLwrangler

OpenTelemetry @opentelemetry.io · 28/09/2026

In just a few days, Amy Super(star) will show us what the Contributor Experience SIG has been up to! 👀 She'll be sharing some of the research the SIG has been working on, along with tips on how to get your foot in the door and start contributing to OTel. Sign up now: buff.ly/6eCrCkv

041

Reposted by YAMLwrangler

Kubernetes @kubernetes.io · 28/09/2026

Hello, Sofia! 👋 KCD Sofia 2026 is almost here! 🎉 On September 29, the cloud native community comes together at the Sofia Event Center. Who's joining? Check out the agenda: kcd.bg/agenda/ #KCDSofia #Kubernetes #CloudNative

kcd.bg

Agenda · KCD Sofia 2026

The full KCD Sofia 2026 program: conference talks, lightning talks, and speakers on stage 29 September at the Sofia Event Center.

043

Reposted by YAMLwrangler

Cloud Native Computing Foundation (CNCF) @cncf.io · 28/09/2026

Keynotes from KubeCon + CloudNativeCon China are now live on YouTube! See how open source tools power AI agents and scale AI training with Kubernetes and KEDA. Watch the full playlist: youtube.com/playlist?list=PLGcxfZT6…

023

YAMLwrangler @jmm86.bsky.social · 28/09/2026

Things about AI agents in production I wish someone had told me before the first outage. 🧵

100

YAMLwrangler @jmm86.bsky.social · 27/09/2026

Materialized views are often overlooked. By pre-computing complex aggregations, they speed up read-heavy queries with a simple table scan, up to 50% faster. Materialized views can significantly reduce query execution time, for instance. #databaseoptimization #sqlperformanceboost

010

YAMLwrangler @jmm86.bsky.social · 27/09/2026

Chaos engineering in CI/CD, like pausing microservices to test system stability, shifts from reacting to failures to proactively validating resilience. Netflix uses it to ensure service reliability. #devops #resilienceengineering

000

YAMLwrangler @jmm86.bsky.social · 27/09/2026

Image update automation in Flux or ArgoCD is like magic until a commit for a new version breaks production. Pinning image tags in prod prevents such magic mishaps. A new version tag in dev was pushed to prod, breaking services. Pinning tags avoids this. #DevOps #ContinuousDelivery

000

YAMLwrangler @jmm86.bsky.social · 27/09/2026

Distributed tracing looks promising until you realize 80% of your services don't emit spans. Start there with 20 services for a solid foundation. Less overhead, more meaningful insights. #DevOps #Microservices

010

Reposted by YAMLwrangler

Cloud Native Computing Foundation (CNCF) @cncf.io · 27/09/2026

Showing up, sharing knowledge, and connecting people is how open source grows. CNCF Ambassador Leon Nunes shares how non-code contributions, working groups like IoT Edge and Wasm, and face-to-face connections power the cloud native ecosystem. Join the community in person: bit.ly/2Id7mzQ

021

YAMLwrangler @jmm86.bsky.social · 27/09/2026

Nexspence v2.9.0 is out: artifacts promote automatically on publish, gated by the scan severities you choose; write-once policies stop released versions from being overwritten; scheduled backups; per-repo "Set Me Up" snippets and a light theme. AGPLv3. nexspence.com #DevOps #DevSecOps #OpenSource

Nexspence Browse view: Docker repository tree with the component details panel open

Nexspence Cleanup Policies list with per-format retention rules for Docker, Maven, npm, PyPI and raw

030

YAMLwrangler @jmm86.bsky.social · 27/09/2026

What running DevSecOps in production actually taught me, beyond the conference talks. 🧵

100

YAMLwrangler @jmm86.bsky.social · 26/09/2026

Our AI agent excels with one user prompt but fails with others, proving robustness to varied inputs is key for success. A weather report generator now struggles with extreme weather scenarios. #AIImprovement #RobustDesign

100

YAMLwrangler @jmm86.bsky.social · 26/09/2026

Supply chain security starts with knowing what you're shipping. Most teams can't list all third-party dependencies in their prod containers. For example, a tech firm shipping a web app had 20 unknown dependencies, exposing security risks. Knowing all components is key. #SecurityFirst #Su...

000

YAMLwrangler @jmm86.bsky.social · 26/09/2026

MTTD improvement has more room than MTTR for many teams. MTTR is visible, but MTTD is hidden until critical. For instance, network outages. #improvement #cybersecurity

100

YAMLwrangler @jmm86.bsky.social · 26/09/2026

Nexspence v2.8.0 is out: Hugging Face Hub becomes the 18th artifact format, Azure Blob Storage joins as a storage backend, and OIDC role sync reads Google Workspace groups. Plus a repo-type filter in the UI and fixes so SSO/LDAP no longer wipe roles. AGPLv3. nexspence.com #MLOps #DevOps #OpenSource

Nexspence repositories list with format badges and a format filter

Nexspence System Admin blob stores tab showing local and S3 stores with usage and quotas

Nexspence Security roles list with built-in and custom roles

020

YAMLwrangler @jmm86.bsky.social · 26/09/2026

Felt like churn until a 4.7.8 OpenSSL update saved our project from a critical CVE, now dependency updates are less stressful. #DevOps #SecurityUpdates

000

Reposted by YAMLwrangler

Cloud Native Computing Foundation (CNCF) @cncf.io · 26/09/2026

Running 1,500+ GPUs and 200+ custom accelerators across 5 hybrid Kubernetes clusters requires reliable networking. Preferred Networks standardized on Cilium to eliminate tunneling overhead and boost model training performance. www.cncf.io/case-studies/preferred-… #Cilium

021

YAMLwrangler @jmm86.bsky.social · 26/09/2026

Kubernetes is powerful — and routinely misused. Here's what 3 years of production incidents taught me. 🧵

000

YAMLwrangler @jmm86.bsky.social · 25/09/2026

Insight: Enforcing semantic versioning for internal services is key — unchecked 'patch' updates that break APIs erode trust. Example: A recent unchecked patch broke our internal API, causing 20% delay in internal workflows. #DevOps #APIStability

000

Reposted by YAMLwrangler

Cloud Native Computing Foundation (CNCF) @cncf.io · 25/09/2026

Final call to submit your production story for the CNCF End User Case Study at KubeCon + CloudNativeCon North America! Submissions close today. Take five minutes to submit your entry. bit.ly/46rC0O5 Be part of our global community at KubeCon + CloudNativeCon NA: bit.ly/3BDI5XL

bit.ly

CNCF End User Case Study and Reference Architecture Contest - KubeCon + CloudNativeCon North America 2026

Take this survey powered by surveymonkey.com. Create your own surveys for free.

011

YAMLwrangler @jmm86.bsky.social · 25/09/2026

New insight: When choosing a Terraform registry module, always verify its source code rather than relying solely on the README. It saved us from a critical bug in production. Check out the examples for better understanding. #terraform #codingbestpractices

000

YAMLwrangler @jmm86.bsky.social · 25/09/2026

Insight: Use Git as production's final say, enforce with admission controllers, not just docs. Example: Netflix limits manual changes to production. Consequence: Minimized errors, ensured consistent app updates. @gitbestpractice @admissioncontrollers

000

YAMLwrangler @jmm86.bsky.social · 25/09/2026

To avoid chaos, implement a comprehensive governance strategy. Testing a network change across AWS, Azure, and Google Cloud shows the unpredictable blast radius. Enterprises must coordinate closely. #cloudcoordination #governancedesign

000

Reposted by YAMLwrangler

KubeFM @kube.fm · 25/09/2026

🗣️ Tanat Lokejaroenlarb, Staff Site Reliability Engineer @ Adevinta, explains how Karpenter fundamentally transformed their Kubernetes upgrade process by decoupling control plane and data plane operations Watch the full episode: ku.bz/T6hDSWYhb

112

Reposted by YAMLwrangler

KubeFM @kube.fm · 25/09/2026

This episode is brought to you by StormForge — automate Kubernetes rightsizing with machine learning. Smarter limits, less waste, better performance. ku.bz/X7ls6SKmr

001

Reposted by YAMLwrangler

Cloud Native Computing Foundation (CNCF) @cncf.io · 25/09/2026

Looking to navigate your Kubernetes certification journey or solve complex cluster challenges? Connect directly with Kubestronauts and CNCF Ambassadors at KubeCon + CloudNativeCon. Register: events.linuxfoundation.org/kubecon-…

011

YAMLwrangler @jmm86.bsky.social · 25/09/2026

The terraform workflow has 4 steps that teams skip at their peril. A thread:

100

Reposted by YAMLwrangler

Cloud Native Computing Foundation (CNCF) @cncf.io · 24/09/2026

Want to share your team's production journey on stage at KubeCon + CloudNativeCon North America? Submissions for the CNCF End User Case Study close soon. Showcase your architecture to the community. Submit here: bit.ly/46rC0O5 Join us at KubeCon + CloudNativeCon NA: bit.ly/3BDI5XL

042