choke_this on Nostr: Bricked my node with 10 channels and 25M sats yesterday. Long story short, I wanted ...
Bricked my node with 10 channels and 25M sats yesterday.
Long story short, I wanted to upgrade my pi4 running Umbrel to a desktop with more power and a 2T drive. Installed Umbrel OS thinking I could successfully migrate to the new hardware without having to close channels.
Well…it worked for about 6 hours, then the “new” node started becoming slow and unresponsive. Rebooting it and none of the docker apps would launch. I was getting errors as if the file system was corrupted.
Frustrated and losing patience I started over…migrating my old node sad to the new. Which was stupid and dangerous since it included all my older channel states. (The channel peers could get all the bitcoin in the channel as a penalty for me broadcasting the old state if I initiated a force close). Lucky for me it only resulted in my 9/10 of my peers force closing so I have all but about 4M sats back on-chain.
Lessons learned:
1. the pi4 was doing just fine, I should have let it go - but I wanted more robust hardware in case fees go up and I didn’t want to open new channels
2. As Darth Coin says, don’t make any dumb, quick decisions if your node goes down. Stop and think on it
3. I really don’t need so many channels or as much capacity, and I’ll probably sleep better that way (but I almost never had failed payments)
4. Test new hardware and OS for a week before YOLOing all your fat channels into it. (I am gonna try StartOS since Umbrel was a bust)
Long story short, I wanted to upgrade my pi4 running Umbrel to a desktop with more power and a 2T drive. Installed Umbrel OS thinking I could successfully migrate to the new hardware without having to close channels.
Well…it worked for about 6 hours, then the “new” node started becoming slow and unresponsive. Rebooting it and none of the docker apps would launch. I was getting errors as if the file system was corrupted.
Frustrated and losing patience I started over…migrating my old node sad to the new. Which was stupid and dangerous since it included all my older channel states. (The channel peers could get all the bitcoin in the channel as a penalty for me broadcasting the old state if I initiated a force close). Lucky for me it only resulted in my 9/10 of my peers force closing so I have all but about 4M sats back on-chain.
Lessons learned:
1. the pi4 was doing just fine, I should have let it go - but I wanted more robust hardware in case fees go up and I didn’t want to open new channels
2. As Darth Coin says, don’t make any dumb, quick decisions if your node goes down. Stop and think on it
3. I really don’t need so many channels or as much capacity, and I’ll probably sleep better that way (but I almost never had failed payments)
4. Test new hardware and OS for a week before YOLOing all your fat channels into it. (I am gonna try StartOS since Umbrel was a bust)