DevOps
Consumer Lag Is Not a Throughput Problem Until You Prove It Is
Lag climbing on one partition while the rest keep up tells you something specific. Here is how to read the numbers before you add consumers that cannot help.
JavaScript
The Dashboard That Got Slower the Longer You Left It Open
A single page app that was fine for an hour and unusable by the afternoon. Three wrong hypotheses, then a heap snapshot showed forty thousand detached DOM nodes.
DevOps
Your systemd Service Will Not Start and the Logs Say Nothing Useful
status=203/EXEC, status=217/USER, and the other exit codes that mean something specific. Here is how to read them and what each one actually points at.
MicroVMs
The Test That Only Failed When It Ran Second
A suite that passed alone and failed in the pipeline. The bug was not in the test, it was in what the previous job left behind on the runner.
Node.js
EMFILE: Too Many Open Files, and Why Raising the Limit Is Usually Wrong
The quick fix is raising ulimit. The real question is what is holding thousands of descriptors open and never closing them.
Opinion
Five Things I Was Confidently Wrong About
Positions I argued for, sometimes loudly, that turned out to be mistaken. What changed my mind in each case.
Shopify
The Shopify Theme Debugging Toolkit
You cannot attach a debugger to Liquid. Here is the set of techniques that actually works for finding out why a theme is doing something unexpected.
Shopify
Migrating a Shopify Store to Online Store 2.0 Without Breaking Everything
A practical migration order, the things that silently break, and how to verify you did not lose content along the way.
Shopify
Shopify Inventory Says In Stock and Checkout Says Otherwise
The storefront cached availability, the customer added to cart, and checkout disagreed. Here is where the state diverges and what to do about it.
Debugging
Your Error Messages Are a User Interface and Most of Them Are Terrible
An error message is read by a person under stress who needs to make a decision. Almost nobody designs them that way.
Bug Hunt
Chasing a Race Condition That Disappeared Every Time I Added a Log Line
The classic heisenbug. Here is why observation changed the outcome, and the three techniques that found it anyway.
Bug Hunt
The Outage Caused by a Successful Health Check
Every replica restarted at once during a brief database blip. The health check was working exactly as designed, and the design was wrong.