The first step of simplification (exiting legacy), be it systems, processes, products, is to know who/what is still using the old thing. Are the users internal (eg admins), external (eg customers) or are the transactions M2M (machine-to-machine)? If M2M, which machines and what are the transactions? This information becomes important for change management (eg transition collaboration and notifications) leading up to the exit.
We’ve spoken many times about how important subtraction projects are to reduce the OSS pyramid of pain.
One of the biggest challenges in exiting an OSS system, workflow or the products / services flowing through them is actually figuring out how they’re used and by whom. The reason it’s a challenge is because there is often no mechanism for tracking use, especially for workflows / services that traverse multiple domains. Even more challenging are the machine-to-machine interactions.
I’ve seen an example where a system no longer seemed to be used by anyone, the last person to have admin rights was made redundant years prior, no documentation could be found that helped and all significant expertise left the building with Elvis. It seemed like the perfect system to shut down and move on with a new workflow. It was shut down and all hell broke loose. Nobody was aware that it was acting as an aggregator and augmentator via M2M transactions. With a lack of expertise, it took days to reinstate. Ooops.
BTW. That wasn’t one of the many mistakes I’ve made in OSS. It was done by a group working on a project adjacent to mine.