SFTP stopped working after you recreated your server

1. Go to My Servers, select the rebuilt server, and open the Files card.

2. Click the SFTP chip in the card's title bar to open the Connect via SFTP window.

3. Use the Copy button on the Username row and paste that value into your client, replacing the whole thing (the part before the dot usually looks unchanged, the part after it is what moved).

4. While you are there, check the Host row too. A rebuilt server can land in a different region, and each region has its own address, like sftp-eu-west.haptic.gg. The Port is still 2022.

To undo this, nothing is destroyed: your old saved session is just a text entry in your client, and you can keep a copy of it until the new one connects.

If you connect a database tool through the SSH tunnel, the same username lives in the tunnel half of the connection, not the database half. The haptic_cfx database login is unaffected by a rebuild. See Connect to your server's database.

Because your password was never wrong. We check your password first and only then look up the server code, so a correct password with a retired code still ends in a refusal, and it looks identical to a bad password from inside your client.

Rotating your SFTP password under Account > Settings > SFTP credentials does do one real thing: it clears an account lockout. It cannot clear the per-address block described below.

We count refused sign-ins per internet connection, because that is how we keep password-guessing bots off the file service. Six refused sign-ins from the same address within 15 minutes block that address for one hour. A saved session that retries on its own reaches six quickly, and some clients try more than one login method per attempt.

While the block is on, we close the connection before the login even starts, so your client stops complaining about the password and starts reporting that it cannot connect or that the connection closed. Two honest details:

• Only time clears it. There is no button for this one, and the hour runs from the last refusal. Turn off auto-reconnect in your client while you wait, because a client hammering the door can push the hour back out.

• It follows the address, not the account. Everyone on your home connection is refused for that hour, and a different network (a phone hotspot, for example) is not blocked. The dashboard itself is untouched, so the file browser on the Files card keeps working the whole time.

A separate amber SFTP locked out banner can appear under Account > Settings > SFTP credentials after a much larger run of failures. That one has a Reset lockout button and clears immediately.

Fix the username first, then connect once. One clean sign-in clears your address's score.

• Resetting or rotating your SFTP password. The password was fine, and the reset does not lift the hour.

• Changing your Haptic account password. It is a separate credential from SFTP entirely.

• Retrying with the old username. Every retry is another refusal on your score.

• Adding an SSH key instead. A key resolves the same username, so the retired server code still fails.

• Restoring a backup onto the new server. It brings your world back, not the old server's code.

If a freshly copied Username still fails after the hour has passed, ask us on Discord or email support@haptic.gg. Include the part of the username after the dot, your server's region, and the exact message your client shows.