
Why does Rust plugin progress disappear after a restart?
A plugin can show new progress during play before writing it to storage. Create one unmistakable record on a disposable world and follow it across a clean stop and start. That separates a missing write from a startup that reads elsewhere.
Follow one change on a fresh test copy
Choose a harmless value such as a test account's balance or a test tally. Use the plugin author's instructions to identify its data file or isolated database. Oxide configuration and plugin data are separate.
- Record the initial in-game value and read the corresponding test data if the tool exposes it.
- Make one player action that changes the value. Record the new value and the action's time.
- Wait for the plugin's documented save moment if it has one, then stop the server cleanly. Read the test file or store again and note any console error.
- Start the same test instance with the same plugin version and read the value through the player's action. Keep the preceding startup log.
Find which step lost the progress
Follow the installed plugin’s own save timing. A file that updates only at a scheduled save may look unchanged during the first part of the test.
| Observation | Check next |
|---|---|
| The value never reached storage by clean shutdown | Read the plugin's save instructions and look for its first storage error. Confirm it can write to the expected location or reach the test database. |
| The value reached storage and startup shows an older value | Compare the active game folder and plugin storage setting on startup. Check whether the plugin logged a format or schema error. |
| The clean check passes and the reported loss followed an abrupt stop | Show the maintainer when the last documented save occurred and whether that interruption could explain the gap. |
Test a second value after the repair
Use the author's supported fix on the same isolated instance and repeat the sequence with a new value. A pass means the written record and the ordinary player's reading agree after a clean restart. Leave the live data untouched while the reason is uncertain.