cloud / systems
Lessons from
the terminal.
Production stories, technical deep dives, and selected notes preserved from the original site—with enough context to remain useful today.
This website was built with prompts
From a broken WordPress site to a nine-language static platform: what building my own website through prompts actually felt like, including the US$15 lesson.
Read the full story →Rebuild.
Observe.
What Real User Monitoring Can—and Cannot—Tell Me
Browser monitoring can complement server-side traffic data, but its sampling, privacy and cost limits matter.
Read note →Manage your website by phone
Building website features from your phone is possible. Here's the honest version: prompts, pull requests, waiting, fixes and the real cost trade-off.
Read note →When Automation Becomes the Bottleneck: Lessons Learned
Nine languages and safe deployments brought complexity. Here is what our recent failures taught us, the CI improvements underway, and how we plan to measure them.
Read note →Codex credits exhausted? How I kept developing my website with ChatGPT and GitHub
A practical fallback: connect a private repository, use ChatGPT to propose changes as pull requests, test them and deploy separately—without buying more Codex credits.
Read note →Why I walk while I work
From walking through meetings to a daily routine: why a walking pad fits my working life.
Read note →A small status line for a clearer head
Which model is working, how much context is in use, and what is this costing? A quiet status line can bring all three into view.
Read note →How I accidentally became a hamster expert in 1998
I built a student homepage. I needed content. I had a book about dwarf hamsters. The search engines took it from there.
Read note →Build an honest ETA display for your coding agent
A tested Python example for reporting the next attention window, input needs and expired estimates, with setup and rollback.
Read note →My coding agent is working. What should I do meanwhile?
Wait, switch tasks or start another topic? How uncertain agent runtimes and interruptions complicate time planning in agentic coding.
Read note →From a notebook to Kraftwerk: why I built my own training app
Too expensive, too many ads, or simply the wrong fit. My search for a training log ended with a notebook—and then an app of my own.
Read note →EFS Bursting vs. Elastic Throughput: How to Choose
Choose an EFS throughput mode from workload shape, predictability and operational risk—not from the default alone.
Read note →How to Monitor EFS Burst Credits with CloudWatch
The metrics, alarms and dashboard relationships that reveal an EFS throughput deficit before it becomes an outage.
Read note →Why EFS Performance Can Degrade After Months
A practical explanation of the long delay between an unsustainable EFS workload and the day users finally notice it.
Read note →What I Learned in October 2026: Making Secure Development Routine
Practical lessons from combining security linting, Dependabot alerts and automated quality gates into a sustainable secure-development routine.
Read note →A Static Site with a Pulse: Privacy-Safe Visitor Analytics on AWS
How CloudFront logs become an anonymous public world map through a scheduled, least-privileged pipeline—without turning the static site into a dynamic application.
Read note →From WordPress to an AI-Assisted Static Site: Rebuilding Without Losing the Past
A practical migration story about preserving years of content while replacing CMS complexity with a secure, observable static platform.
Read note →The Six-Month Time Bomb: When EFS Ran Out of Burst Credits
How temporary EFS burst capacity concealed an unsustainable workload for six months—and what the incident taught me about monitoring slow-moving failure.
Read note →Past and Future
A personal look back at the systems journey so far and the direction ahead.
Read note →My vita (pdf)
The archived curriculum vitae from the original website.
Read note →check-mk and snmp
Practical notes on extending checkmk monitoring through SNMP.
Read note →AWS Autoscaling / ELB Health Checks
Historical implementation notes on Auto Scaling and load-balancer health checks.
Read note →Ubuntu 18.04 (AWS) ext4 FS
A field note on working with an ext4 file system on Ubuntu in AWS.
Read note →Analyze apache logs with awk
Small command-line techniques for turning Apache access logs into useful answers.
Read note →HTTP Pragmas
A concise archive note on HTTP pragmas and cache-related headers.
Read note →Install rpm on Ubuntu
Historical notes on handling RPM packages from an Ubuntu system.
Read note →AWS: restart EC2 on failure
A 2019 recovery experiment using Route 53, CloudWatch, SNS and Lambda to restart a single EC2 instance.
Read note →Temporary E-Mail inbox
A small utility note from the original site about temporary email workflows.
Read note →AWS::EC2 how to enable custom metrics
Historical guidance on enabling custom EC2 metrics and improving visibility.
Read note →wordpress and let’s encrypt
A practical archive note about protecting a WordPress site with Let’s Encrypt.
Read note →Sat my AWS SysOps Admin exam, April 10th 2019
A milestone from 2019: passing the AWS SysOps Administrator exam.
Read note →