Made the Docker automation reproducible and instantly recoverable (Phase 3)

Building on the earlier move into Docker, the automation that quietly runs the business now runs from versioned, labelled images with its working data kept in dedicated, snapshotted storage – so any run can be traced to an exact version, and the whole setup can be rolled back to a known-good point in seconds without […]

In this series

On this page

Free Revenue Lifecycle Assessment

Connect with Marissa Wright to receive a free Revenue Lifecycle Assessment Report on your own business.

Book a consult →

Building on the earlier move into Docker, the automation that quietly runs the business now runs from versioned, labelled images with its working data kept in dedicated, snapshotted storage – so any run can be traced to an exact version, and the whole setup can be rolled back to a known-good point in seconds without losing data.

  • Froze the automation's code into versioned, labelled container images – every scheduled run is now tied to an exact, reproducible version, and rolling back becomes as simple as re-running a previous label instead of un-picking changes by hand.
  • Moved the automation's working memory – the week-over-week SEO history, the grant-notification subscriber list, and each job's last-run markers – off the Mac and into dedicated storage volumes, so the data no longer depends on the machine's own folders.
  • Added automatic daily snapshots of that data, kept as a rolling set of recent and weekly backups (held locally, never in the cloud), giving a clean recovery point if anything is ever lost or corrupted.
  • Rollback now has two independent levers – restore an earlier code version, or restore an earlier data snapshot – and, true to the earlier phases, the previous setup stays in place as a fallback so nothing is a one-way door.