Jeremy Stanley writes:
It's all stored in a relational database,
This isn't quite true:
- order of shutdown matters because you want to be sure queues are flushed
- several queues (bad, shunt, queues internal to archiver plugins, specifically HyperKitty) do not get flushed (bad cannot be flushed)
- some database items (eg, custom templates) get cached on disk, and people sometimes modify them on disk and/or add new template directory to the template search path
which you should be able to extract with a tool like mysqldump and then later source that same dump into a new empty database.
This is often unnecessary. If you're going to be running the new Mailman on the same host, you can probably keep the old database running, and point the new Mailman applications to it.
You can find some information on database setup in the documentation here: https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/docs/databa... Some familiarity with backing up, restoring and manipulating databases is a big help for these sorts of tasks, otherwise it may have a fairly steep learning curve.
Let me second Jeremy's advice above. Consider the skills you have available to you when deciding how to do the migration, and feel free to discuss your plans here.
-- GNU Mailman consultant (installation, migration, customization) Sirius Open Source https://www.siriusopensource.com/ Software systems consulting in Europe, North America, and Japan