Cerner’s Confluence system was running on 3.x which was notorious for being hard to upgrade to newer versions of Confluence. We knew going in that this would be a challenge, but in the initial two days we spent at Cerner’s office in Kansas City, we weren’t able to delve deeply into the customizations Cerner had made. Over the next four months, we tried to work with app vendors and Atlassian, but no one would support Cerner’s hacked versions so we figured it out ourselves and successfully delivered a runbook to build an upgraded system that worked similarly but better than their old and hacked setup.

This system was running on Linux servers with an Oracle database. The company had significantly modified core application files, the database, and had written numerous custom add-ons. They had also modified some third party add-ons. All of this contributed to tremendous complexity in the upgrade.

As the project manager, I performed the initial discovery to uncover as much as possible about the original system and its infrastructure. I proposed the schedule, which required collaboration between Adaptavist and the company’s team involved with this project, and I maintained all documentation including schedule, regular updates, meeting notes, risk log, and invoicing. I also worked closely with the company’s Atlassian Technical Account Manager (TAM) to ensure we were providing the best support possible backed up by Atlassian’s internal engineering experts.

The company’s project team elected to perform the final upgrades themselves, so a work plan was developed using their development server, and they were able to use the work plan to successfully upgrade their test server. We delivered this work plan ahead of schedule, and overall were under-budget.