We are using Confluence for IT documentation and Jira for our IT project management.
We have created a unified hierarchy that we use in Jira by having 3 main categories (Infrastructure, Operations, and Services/Programs) that make up our 3 projects. We then break each one out using components giving us a relatively simple way to organize stories/tasks for the various things we manage. This has worked out pretty well and can tie into creating simple reports of related issues within confluence.
On our server instances we were breaking out the confluence spaces that match these at the component level, which made for a lot of spaces, but they were relatively smaller which made pdf export possible so that we can have an emergency copy of the documentation as part of our disaster recovery plan. We used the refined app to create a site map to make navigation easier.
Now moving to confluence cloud we're looking at how and if we need to re-organize things. The navigation of a lot of spaces like we have is a bit clunky, and the refined app involves creating a whole different website that doesn't support all our macros, so we're not sure that's a direction we want to stick with.
We've considered creating just 3 spaces to match the 3 projects, which would sort of follow some of the best practice guides we've found on creating a space for each "project". However this will create HUGE spaces that would be 1000+ pages. We have one space for our core infrastructure (so there are other components still broken into other spaces) that is around 600 pages, doing a pdf export of it in cloud took 20 minutes and created a pdf over 1 GB (it also took a long time on the server instance), opening a 1 GB+ pdf can be problematic, and if we merge all our spaces to 1 larger space with a parent page for each component it will get even bigger and become a bigger and bigger issue.
We realize we can create custom exports of individual trees, but at least on the confluence server side, we can't do that in an automated fashion. We create a pdf export using the confluence CLI add-on weekly via an automated task so that we always have a current export. If there's a way to do custom exports through the rest api or the confluence cli api then maybe we can figure out the giant spaces.
So we're hoping we're not the first people with this problem and someone can share some insight on how to break out the documentation in a way that's not too cumbersome to navigate and that will scale into a backup/disaster recovery solution with some ease.
