Writing a Production Cutover Runbook That Survives the Night
A minute-by-minute cutover runbook template with named owners, a point-of-no-return gate, curl and psql smoke tests, and a rollback you rehearse rather than hope for.
Growth & Migration
SAP ECC to S/4HANA: Automating Custom-Code Remediation Analysis
The real risk in an SAP ECC-to-S/4HANA migration hides in years of custom Z-code. Automated scans and usage data let you size the remediation with evidence before the 2027 maintenance cliff.
Don’t forget you are out of extensions on Server 2012 — there is no year four
Extended Security Updates for Windows Server 2012 and 2012 R2 end on 13 October 2026. Three years was the whole runway — there is no year four to buy. Find your last box before the date, not after it.
Ubuntu 22.04 to 24.04 In-Place with do-release-upgrade
An in-place Ubuntu 22.04 to 24.04 upgrade works, but only if you prep: snapshot first, disable third-party PPAs, handle the config prompts deliberately, and keep a tested way back.
PostgreSQL 15 to 17 Major Upgrade with pg_upgrade
A rehearsed in-place PostgreSQL 15 to 17 upgrade with pg_upgrade –link: pre-flight checks, matching extensions, statistics rebuild and a tested rollback.
This Fixes: pg_upgrade ‘could not load library’ Extension Mismatch
pg_upgrade aborts with ‘could not load library’ when an extension has no matching binary in the new cluster. Install the target-version contrib and PostGIS packages, re-check, then upgrade.
Migrating a Fleet From SSH Passwords to Key-Only Auth Without Locking Anyone Out
Switching off SSH password auth before keys are proven is how you brick a fleet. The safe order: distribute keys, verify every login on every host, keep a break-glass path, then disable passwords last.
Upgrading a Service From .NET 6 to .NET 8
How to retarget a real ASP.NET service from .NET 6 to .NET 8 LTS: the TFM and global.json bump, the analyzer, and the breaking changes that actually bite.
Process Design Before Tooling — Always
Automating a broken process makes the business bad at the process faster. This is one of the oldest observations in operations management, and technology teams rediscover it painfully, at real cost, every time a new tooling category makes automation newly tempting — which is, currently, constantly. I’ve written about the data model and the architectural …
Your Data Model Decides Whether Your CRM Works
Two organisations can buy the identical CRM platform, configured by equally competent implementation teams, and end up with wildly different results — because the platform was never the variable that actually determined success. The data model underneath it was, and almost nobody scrutinises that model with anything like the attention the platform selection itself receives. …