Bob, i hope this isn't the last time you go down this rabbit hole!
It won't be. Plan to revisit in at least another 18 months.
#LNMETA
Some more thoughts on being a top 250 LND node op... Again...
- This was a good learning experience, compared to my first experience of needing to hand-roll all the LND/c-lightning stuff from source, and compile bitcoind with zmq support. Throwing A LOT more liquidity at it this time really helped me get a sense of what it takes to be a big boi on the Lightning Network
- Learned more about the "quantum" state of liquidity. Eg: LOOP experiments draining my liquidity, putting in a "quantum state", where I can't regain that bound liquidity for use, until I wrap everything up via "lncli closeallchannels" sorta deal
- Avoid LOOP. It's a liquidity drain, and pure profit chasing - Opened 3 wumbo channels with LOOP before I learned my lesson of avoiding them entirely. Was pissed off all that liquidity was lost to me in "quantum space"
- Many nodes are parasites (like LOOP) - They will drain all your liquidity to their side, and nothing flows back your way in some cases (ln.nicehash), to even get you back into organic balance. LOOP being the most egregious, because they will just close your channel on you when they've sucked enough liquidity out of you, and don't even have the fucking decency to leave the channel open, for you to recoup routes back your way. Pure vampirism and it disgusts me.
- Was very impressed with Umbrel. Was a bit disappointed to learn about the apparent background drama of Umbrel pretty much being a clone of another popular distro (RaspiBlitz?) but they've really added a good layer of polish onto node distros, and love the ability to reset your node back to it's "last good known state" just by entering your Umbrel seed words.
- With the increased popularity of Umbrel, and other very well maintained node distros, Lightning looks like it's really shaping up well in terms of "now everyone can run a server" - It's as simple as getting a Raspi4, a 2TB SSD drive, half a brain, and you can start having fun dipping your toes into Lightning, pretty safely and cheaply.
- Rebalancing: Absolute nightmare. After further consideration, with Rick having looked over the scripts, and understanding the nature of what it's doing and the problems that need to be solved, it's not an easy problem, mostly owing to the constantly fluctuating state of the network.
Every channel you are trying to rebalance has to get tested across MANY routes. Some of them good, some of them bad. Due to a function of how the LN works, you *HAVE* to check every node, on every route, every time - and there are A LOT of bad nodes out there that are not responsive, and the rebalancers don't remember the state of which nodes are bad, when balancing from one channel to the next one, because - well, the nodes could have popped back up again since you last checked - Needs more logic for how to handle "bad" peers organically. The onus is on the person running the script, to exclude nodes to check, and when you have 100+ channels, fuck that hand-management reconciliation noise.
- I was definitely chasing profits. This is what was leading to the stress in my life, overall, and it was starting to impact me in meatspace pretty badly. I'm not satisfied "just leaving" something to run. If I'm going to run a node, I'm going to be an active manager, and look at this like getting into mining for profit. Don't have any interest in passively running a LN node. I'm either going to do it right, or not at all. Chasing LN node profits was more frustrating than running a mining farm - A comparative cakewalk.
- Meatspace is becoming increasingly complex for me for a little bit, and running the node was becoming negative energy, that I am desperately seeking to rid my life of, at this stage.
- Torque basically summed it up best
