pull down to refresh

Over the past two years I ran an amateur routing node that was in the top 300 on lightning terminal.
I have decided to stop for the following reasons
  1. Profitability
    I routed many payments but was never profitable. My total lifetime node statistics are in the table below:
    Line ItemLifetime
    Payments Routed24,916
    Value Routed944,239,099
    Revenue Earned14,790 [15]
    Onchain Costs784,204
    Offchain Costs9,665 [474]
    Percent Cost5,367%
    Profits-779,079 [-825]
    Could I have done a better job managing my expenses and routing fees? Yes. There are lots of people who run lightning nodes profitability. I was just not one of them. If I had spent more time optimizing my fees and I raised my fee rates I could have been.
    But even doing that wouldn't stop random force/regular closures which require you to reallocate your capital to another channel (which means you have to spend sats on opening fees).
    If you want to run a routing node successfully you really need to treat it like a business. And there are actually a lot of expenses in doing so that aren't apparent upfront.
  2. Responsibility
    I am happy that it is possible for anyone to run a routing node. However just because you can doesn't mean you should.
    It's a big misconception is that everyone needs to be a routing node. While on paper most people can easily set one up. The average person is much better off connecting to an existing, reliable routing node than trying to make their own. Managing inbound/outbound liquidity requires some effort. And if you don't keep up with it, you might be doing more harm than good to the network. Over-leveraged inbound and outbound liquidity results in more failed payments on the network.
    I had about double to triple inbound liquidity so I saw a lot of failed transactions and I probably contributed to an overall worse UX for the rest of the network because of those failed transactions. I did not run my node very responsibly because I was doing it as a hobby.
    As silly as it sounds, I just really wanted a higher lightning terminal score and larger node appearance so I bought inbound liquidity to make my node look larger.
    So while I did have a large amount of outbound. It really was a lot less than what it appeared to be. And while it was fun for me to have a larger score, it probably wasn't fun for the people not able to route their payments.
  3. Risk It is risky to run a lightning node. The software is still experimental, bad backups can lose all your funds, all your money is in a hot wallet. Things can go wrong!! Am I saying don't do it? No, but what I am saying is only put in what you're willing to lose. Because bugs, backup mistakes, software/physical security breaches can happen.
    I am willing to accept these risk but only with less capital than I had allocated when I was a routing node. I only had success forwarding when I started to have many channels with large bandwidth. It's just no longer worth it to me anymore to maintain that.
So going forward, I am only going to use private channels and focus on my usage not forwarding payments for others.
90% of wannabe routers are just playing a PvP MMO where they steal from each other while pretending they're all helping Bitcoin together.
On a long enough timeline, unfit nodes either give up or become profitable/useful.
The best routers are also merchant/service nodes and routing profit is just byproduct of having a large active node that people want to pay (for goods and services).
This is how it ought to be! Nodes specialized in routing alone are just trying to be middlemen/rent seekers.
reply
Can you elaborate in more detail on what you mean by "stealing" from each other.
reply
Stealing implies non-consent. What OP did was presumably with consent because he decided what to charge for routes on his channels and he approved every spend that incurred fees. So was it actually stealing? Not really, but the outcome is the same, OP has fewer sats and nothing to show for it other than experience. Many node setups don't require a human to approve every step so these are especially vulnerable to wallet draining "attacks". But the operator presumably consented to running the automation in the first place.
reply
Ok, thanks for explaining what you meant. I really think we hurt ourselves and dilute the impact of words like steal when we use them incorrectly. To turn it around, how does a node op using a low fee node to re-balance know WHY the node operator has a 1 sat fee set? It is easy to imply malice in others and virtue in ourselves. Its really foolish. Neither node operator can read the other's mind. If you offer anything a price below cost people will take it from you. Its a great deal! We all like to save money but when we are on the other end we don't like it that much.
reply
Fo sho, Like if I push a button and I get a sat, I'll keep pressing the button. Then someone proves to me that I'm actually taking sats from their node and they do not consent to me "stealing". I'm a good guy so I stop pressing the button, but really they should harden their node so the button can't "steal" in the first place.
reply
Yeah stealing is not the right word but screwing other nodes over to balance their own liquidity (splicing improves this by the way)
reply
Ok, how are people being screwed? Not following what is being done to who or how. If a pleb node operator has low fee rates / fees set and another node(bad guy node) is re-balancing using a channel with this pleb how is this screwing or stealing? I'm not saying it is or isn't but I don't get what is being said here.
reply
I mean its all within the rules so how are they being screwed right? ¯_(ツ)_/¯
People feel as though they are being screwed by it because they charge the same fee regardless of how much liquidity gets used up. A fee that scales to how much liquidity is being used up would make them feel much less screwed over.
Basically, they're expecting to get many fees over many lightning network transactions, but they end up only charging that 1 fee for their entire channel size and it frustrates them lmao.
reply
I would recommend folks go into running a lightning node with the learning mindset not the earning mindset. Adjust your time preference my friends. I do get the frustration though if you go into it thinking this is a way to make easy bitcoin and support lightning. In my experience it is not easy but I'm still learning. Learning is rarely without a cost and IMHO learning more about bitcoin and the tools around it are more undervalued than bitcoin itself.
reply
I see a profit opportunity right now lmao. Pop into the lightning network with a node that has a splicing implementation, let people rebalance through your node and splice in more liquidity. I believe LND was more concerned with creating "Taproot Assets" than they were improving their implementation, so you can probably profit off of LND users in this way.
reply
Perfect. I also had a romanticized idea that each person would have their lightning node routing thousands, if not millions of transactions per month. However, this works neither in a small environment nor in a hyperbitcoinized environment. The number of hops between sender and receiver determines the efficiency and effectiveness of the network. The smaller it is, the better. That's why we will have giant nodes, conglomerates of connections, and small nodes with few connections to these conglomerates. Only then will the network remain stable and active.
reply
Thanks for sharing. You make some valid points to consider. I can see a few reasons for running a routing node.
  1. Profit: This is not easy and you are going to be competing with stronger competition as more enter the market
  2. Altruism: Supporting the network. I don't know how you do this without running a node at a loss
  3. Greater privacy/insight: Thinking of @darthcoin's guide with Blixt. Using private channels with your own node as a routing node.
The way I think about this is to start with what are your goals. I could have mentioned one more which is learning how it all works. When you come to something complex and new like lightning you have to set your expectations. I wonder how many plebs are making an actual profit running a routing node. What I mean is yeah they may show a profit in sats but what about the time ROI?
reply
Using private channels with your own node as a routing node.
This.
private spending node == private channel ==> my own routing node <===> Lightning Network
I'm not really aware of anything on the LN roadmap that can address all the concerns listed here, so for now this will remain my setup.
Private channels to a major routing node are almost certainly a privacy risk, at a minimum that major routing node has metadata about your transactions, but it probably has a lot more than that.
reply
Yeah, I started more with 1,2,3 and now I'm more focused on 3 for sure.
reply
Yeah, the war is coming.
reply
I feel like most people go from routing node to private node but it should be the opposite.
I don't blame anyone though, I wanted to run a routing node too at first. But with some crazy luck, my SSD broke a day before I wanted to open channels, lol
I think I'll stick with a private node, at least for a while.
Good post.
reply
The important thing is that it's possible for anyone to run one, affordably and without permission.
reply
Good take. A routing node provides a service for a fee, therefore operating one is a business, just like mining. This may not be for everyone, either due to the attention, planning and know-how it requires.
reply
Appreciate the sharing. It's not to discourage node runners, but routing is a whole different game. Open channels if you want to be well connected and self reliant. Not because you want to make a profit.
reply
Maybe share how many payments you made too. You could estimate how much you saved in on chain fees.
reply
I had the same experience with my LN node wasted a few sats learning that routing is not for me and if I wanted to make a good go of it, it needs to be coupled with a service, it can't just be me trying to source connections for liquidity but offering my channel capacity as part of services or goods I would be selling
reply
I agree. One of the biggest problems with tutorials teaching you how to run a node is they teach you this with the assumption you want to be a routing node, but there are many reasons to run a non routing node. There are barely any tutorials or guides that go into it.
reply
I noticed that you closed our long lasting channel earlier today and naturally I was heartbroken. 💔
Thank you for your explaination I completely understand your point and I agree that hobby routing nodes might be more harmful than good for the network. Good luck on your future adventures!
reply
I'm about to discontinue the routing here as well. I also run a node for about 2 years and also ran in negative balance. Here's some topics I find relevant:
  • People clever and or richer than me will make it happen profitably
  • TOR connectivity is laggy and some time non responsive
  • risk of being hacked
  • risk of dying while a lot of bitcoins alocated within LN and no relative would be able to retrieve it (close channels and withdraw to sell is too complicated to do even with a tutorial)
  • You can use a mobile wallet and a private channel with some big nodes or specific channels while keeping your private keys without the need of a computer and you can send LN free
  • node restoration is a mess and pricy
I JUST DON'T KNOW why the community still pushes for ppl to host their own node!!
reply
That was basically my approach, to have a node for myself, not a routing one, so if I need to send some sats I use my own node. A profitable Node seems like a very hard task, something that I kind of saw from the very beginning. I did not think about hurting the network... that left me thinking. Thanks for your valuable feedback.
reply
You raise very good points which give me food for thought about my own node. @DarthCoin has expressed similar sentiments in the past. It's a complicated issue. I'm sure you'll get some pushback.
reply
Are these numbers considering the fees closing all current the channels?
reply
This post was featured on This Day in Stacker News.
reply
This is a great article thank you for writing and sharing your story.
It will be interesting to see if the idea of a routing node will survive, or if routing payment profits will only be secondary to another business model.
We run a mainnet node too. I've used it to experiment with routing but now it's more to hook up new e-commerce sites and apps and payment services.
This is a great reminder for me to stay focused!
reply
Thanks for sharing.
reply
Yeah Lightning is a rich man’s game. I was hoping it was going to be a similar technology to layer-1 but I’ve been less than enthusiastic about using the technology.
reply
Running a routing node is most probably not profitable for 90% of the node runners. It requires dedicated hardware, software, automantion and a lot of manual intervention. It is often a full time job.
reply
maybe such solutions will allow to run the node with low effort and still without loss, in some future anyone has deeply tested CLBOSS?
reply
Thank you. So for an average user you don't recommend to run a node like umbrel. Even If I consider myself an advance user, I don't consider myself and advance node administrator. I have no idea at all, neither the resources you mention. This is a huge alert for many. I appreciate your sincerity.
reply
You have me thinking with this post. If one wanted to run a routing node to support plebs and wanted to do zero fee routing. How would you do so without losing your shirt in fees. My guess is that it wouldn't be easy and you'd need to be very selective with whom you open channels.
reply
I came to a similar conclusion. I'm glad that I experimented with running a routing node, but honestly I wasn't very disciplined at maintaining my channel balances and uptime. I figured the network would be better off without my sub-par node, so I wound it down.
Now I view routing in a similar fashion to mining, in that everyone can participate but only the professionals will be profitable. As it should be IMO. I am content in sticking with the likes of Phoenix and Blixt for my lightning needs.
reply
Welcome to the next bull run. Fiat profiteers running amok
reply