There’s a few different options on how to restart your Umbrel node, starting from a simple reboot to effectively deleting everything on your node and starting over.
Simple Restart
If you’re able to log into the Umbrel dashboard, the easiest way is to go to “Settings”->“Restart”
If that option is not available, your next option is to SSH into the machine. Run this entire set of commands once you’re in:
sudo systemctl stop umbrel-startup && docker system prune -f && sudo systemctl start umbrel-startup
This will stop Umbrel processes safely, purge the system of various unused containers/images, and then restart the Umbrel scripts.
Additionally, you can also run:
sudo reboot
This will actually reboot your entire machine, which will reload some other non-Umbrel related things. The only downside is that you lose your uptime record (assuming you care about that).
Restart after a Stalled Upgrade
In rare cases, Umbrel will get stuck on an update. This happens most often when Umbrel is being run on untested hardware/software. If you are able to ssh into the node, run this command to kill the update and restart the umbrel.
rm -f ~/umbrel/statuses/update-in-progress && sudo reboot
Nuke Everything - BACKUP REQUIRED
If your node seems beyond repair, but you don’t want to scrap your entire download of the blockchain, then you can run the following command:
sudo systemctl stop umbrel-startup && sudo rm -rf ~/umbrel/lnd/!(lnd.conf) && sudo rm -f ~/umbrel/db/user.json && sudo rm -f ~/umbrel/db/umbrel-seed/seed && sudo systemctl start umbrel-startup
This will stop everything, delete all your user files, seed phrase, 2FA, password, etc. You’ll be presented with a login as if it’s the first time using Umbrel and you’ll want to restore from seed.
Important note - do not restart or shutdown after you do this, lnd will actually be in recovery mode and doing some work in the background. You do NOT want to interrupt it