Jason Morse

NetApp to Azure Files migration

Storage migration · 2024–2025

We were paying a third-party vendor for cloud storage that had become expensive, and we were already running everything else on Azure. Moving the data to Azure Files meant dropping the contract and consolidating onto one platform. The catch was volume: millions of clinical files that had to be retained for compliance, so none of them could be dropped. Some transfers ran longer than twelve hours, and a failure meant starting that transfer from the beginning.

I wrote Kubernetes jobs that tracked their own progress and picked up where they stopped after a failure. The security review asked for a way back, so I documented rollback procedures and kept the original files in place until the migration was verified.

Production cutover finished inside its maintenance window. The old storage was decommissioned in stages afterward.

Jenkins migration between AKS clusters

Platform recovery · 2025

The team that ran our shared build infrastructure was dissolved. The platform was still critical to release work, and overnight it had no owner.

I planned and ran the migration out of their cluster into ours, adding dedicated node pools in Bicep so build jobs would not compete with application workloads. The new subscription sat behind a strict allowlist, so I had to find and get approval for every external build dependency, mostly by watching which ones failed first.

The migration took just over two months across four environments plus disaster recovery, and ownership went back to the product organization.

Standardizing Azure RBAC

Identity & governance · 2022–2024

Access across the Azure estate had been granted case by case over several years. Least privilege was difficult to enforce and audits took longer than they should have.

Working with engineering and security, I defined a standard set of role groups mapped to actual job functions and built on least privilege, then connected them to our identity governance platform so access could be delegated and reviewed instead of handed out one request at a time. I automated the rollout in Bash and Jenkins.

Per-subscription setup went from more than a day to about 30 minutes. The platform team took the automation into standard provisioning.

Monitoring and availability platform

Monitoring & availability · 2025–2026

I own our monitoring and availability platform. The core of it is a synthetics framework built on Playwright and running in Azure Functions, checking real user journeys across services every sixty seconds from multiple regions.

I also designed and shipped the integration between our paging and ticketing systems, so an alert and its incident record stay in step in both directions without anyone copying fields between them. I wrote the Azure monitoring standards the organization works to.

Availability reporting is now used across several product lines, and alerts close in both systems without manual cleanup.

Log Analytics consolidation

Cloud economics · 2025

Log ingestion spend was climbing because every workspace billed on its own. I consolidated them into a central cluster, which earned a volume discount per workspace and brought production logging spend down by roughly half.

That would not have held on its own. Any new workspace would default back to billing separately, and the savings would erode as the estate grew. I specified the platform change that enrolls new workspaces automatically, and it shipped upstream.

Production logging costs came down by about half, and stay down as new workspaces get added.