Selfhosted

I have been on the same boat a while ago. Of course it was a deployment that caught on and was serving longer than expected.

I don't recall how many versions I skipped exactly (1-2 years worth of updates). Of course no backups set what so ever.

I looked at change log of I need something specific, there were changes in docker compose file. Did my best effort to make it succeed.

My worst case scenario was that I will have to import everything again and made sure I have all my labels, tags and settings backed up.

Nowadays I'll just snapshot the whole VM in Proxmox.

I’ve not had good luck finding something that gives me the confidence to go about it

Now's a good time to make sure you have good backups.
Knowing you can fallback to your backups helps a lot with confidence.

This is how I handle it for most software: Read (or at least skim) the changelogs for all minor and major versions between your current version and the latest version.

If you're using Docker, diff your current docker-compose to the latest one for the project. See if any third-party dependencies (like PostgreSQL, Redis, etc) have breaking changes.

If there's any versions with major breaking changes, upgrade to each one separately (eg. 1.0 to 2.0, then 2.0 to 3.0, etc) rather than jumping immediately to the latest one, as a lot of developers don't sufficiently test upgrading across multiple versions.

Take a snapshot before each upgrade (or if your file system doesn't support snapshots, manually take a backup before each upgrade).

...or just don't read anything, YOLO it, and restore a snapshot if that fails. A lot of software is simple enough that all you need to do is change the version number in docker-compose (if you're using Docker).

I haven't gone through your specific case, but generally what I do when doing a major update with potentially breaking changes:

• Read the upgrade guides, if they have them. Some devs will put them out if they know their users will encounter issues when upgrading. If they don't have an upgrade guide, there might be some in the change logs. Going from 1.17.1 to (assuming) 2.x.y, check the change logs at 2.0.0.
• Backup everything. I'd recommend doing this on a regular basis anyway.
• (If you're running it in a docker container) Setup a second instance, restore the backup, then run the upgrade. You'll be able to check to see if it breaks at all. If it works, you can just destroy the old one and use the new one
• (if you're not running it in a container) with the backup, try upgrading. If it breaks, you should be able to uninstall & reinstall the old version, then restore the backup

Paperless provides an export tool that lets you export all your documents. There's no guarantee that the import will work on a higher version, but usually it's fine. Just make sure that you have a clean install without any documents before importing from an old export.

It's also possible to export all documents and metadata via the REST API, though that's a bit fiddly. Then you can import them the same way. Again, this should only be done into an empty system, because otherwise your object IDs will be all scrambled up.

In these cases I'll do the same thing other people here seem to do as well. Do a backup (or snapshot) and then I'll try to just do it. Obviously read the documentation on updates and major version upgrades first. I think that's fine in the case of paperless-ngx.

Either it works or it doesn't. In that case I'll gather error logs and information for debugging and roll back to the backup. After a successful major upgrade, I often go through the settings and config and check about all the things that have been added or changed in the meantime and make sure they're set to my liking.