I had a quite strange situation. This morning a peer that I know personally notify me that my node what offline for at least 2 days. I tried to connect to Umbrel from the local network and it was not reachable (from GUI, didn’t try SSH at the time). I unpluged the power and replug to force the restart and could connect again but Bitcoin core was not synching. In the past I had a similar problem and it happened that peers.dat and/or settings.json was corrupted due to the improper shutdown. I remove them and rebooted. It start synching again, all seams fine.
By the way, what am I doing when I delete these peers.dat and settings.json file? I ask my node to get all peers information again? and reset some settings to default? which settings?
Then another peers that I know as well, contact me to ask if I did something with my node. I explain the situation and that it should be back to normal but then something strange happened.
The other peer had a problem with his node 5 days before and by mistake it triggered a force close of all his channels. Only some of them did closes, due to the current mempool situation, but our channel as soon as I was back online did finally close. My node did send a transaction with a huge fee (300ksats) without that I did anything explicitly in this way, that apparently push the closing somehow.
Looking at the explorer (mempool.space) the actual closing transaction was send with about 2700 sats if fee, it should have never been included in current mempool situation. But then a CPFP did push it, and this descendant tx had a lot of fee, 300ksats, and did confirm. This transaction come from 2 address, one labeled Lightning Anchor by Mempool.space and end up on one of my address where I receive what looks like my side of the channel (or another channel closed? not sure).
Any idea what could have happened? Is it only due to the other peer? Is it due to my node? both? How can I prevent this to happen again?