Before you upgrade2.x

The upgrade reads your 1.x config, including one from the old ghcr.io/nzbdav/nzbdav image, and imports it into a new database: providers, Arr instances, categories, the SAB API key, WebDAV and admin logins, the mount directory, the library and its item ids, history and the queue.

Your 1.x files are opened read-only and never changed. Still, work on a copy.

Who should not upgrade yet

• DUMB users. 2.x does not ship binaries packaged for DUMB yet. Stay on 1.x for now.
• 1.x on PostgreSQL. The importer only reads a SQLite 1.x database (db.sqlite). Without one, 2.x starts as a new, empty install instead of importing, and import-report says no C# database.
• 1.x older than v1.3.0. Update 1.x and start it once first.
• nzbdav-dev/nzbdav 0.6.4. Migrate to the latest 1.x and start it once first.
• You need a 1.x feature 2.x does not have yet. See What’s new in 2.x.

Steps

1. Stop 1.x so the copy is consistent, then copy its config directory:

   docker compose stop <1.x service>
   cp -a ./config ./config-2.0

2. Optional dry run. This prints what will be carried over, what will not, and why, without importing anything:

   docker run --rm -v ./config-2.0:/config ghcr.io/infinidysk/infinidysk-next:edge import-report

3. Note the time zone 1.x ran in. 2.x needs the same TZ: 1.x stored local times, so a different zone shifts every file date, and media servers may then re-read your whole library.

Next, if 1.x used an rclone sidecar, move off rclone. Otherwise go straight to Run the upgrade.