Lightning node for fees

[vapapgr]

I m new to LN and i decided to run a node for experiment and to get fees. I ve already destroyed an old HDD trying to download the blockchain. So ill buy new equipment . What do you recommend ?  

[joniboini]

What kind of recommendation are you looking for? The hard drive? The network card? The CPU? Or an entire VPS service to run a node remotely?

I've heard some people run LN node on a Raspberry Pi with https://github.com/rootzoll/raspiblitz. You might want to check it out. CMIIW.

[vapapgr]

What kind of recommendation are you looking for? The hard drive? The network card? The CPU? Or an entire VPS service to run a node remotely?

I've heard some people run LN node on a Raspberry Pi with https://github.com/rootzoll/raspiblitz. You might want to check it out. CMIIW.

Seems helpful . Thanks

[Pmalek]

I don't want to discourage you from using the Lightning Network, but if you do it solely to earn on the routing fees expecting to make a decent income, I am afraid you will be disappointed. You won't be able to pay off the purchased hardware by routing payments. Someone correct me If I am wrong, but if you were to route hundreds of thousands of payments, you wouldn't even make $1. Because the money is routed through the cheapest channels. If your route is more expensive than the guy next to you, he will be the one whose channel is likely to be used.

On the other hand, if you want to experiment with off-chain transactions and taking the load off from the main Bitcoin blockchain, that's a great thing to do.

[mocacinno]

Just to give some extra, real-life data from my lightning node that has been running for a long, long time:


Sounds like a lot, right? To avoid any cofusion, that's 97 sat's, not 97k sats

And before you say that i haven't been online that long, i've been running since one of the first versions of c-lightning was available for the testnet... I've reported bugs from the time they advised you only to run on the testnet, even before you could actually run on the main net (iirc).

Running a lightning node for profit.... nah... You'd need tons of heavily funded open channels. It would cost a small fortune to set something like that up, and even then, i doubt you'd break even.
Set up a node to help the lightning network, set up a node to learn, set up a node so you'll always have some open channels to quickly make or pay invoices... But don't expect a steady income from them.

EDIT: a trip down memory lane... There's still some history from my (now defunct) testnet lightning node: https://1ml.com/testnet/node/03a94943f6645e5cd958ab5d8d3bdf19116ba38e07f22411fb041bca1124e9470d/history

[Husires]

@Alex Bosworth report on  Feb 19, 2019 that his node routing more than $10,000/month and getting paid 0.25% for routing
if you buy hardware for $150 and routing more than $10,000/month you can have ROI after 4 months (lucky case)

https://twitter.com/alexbosworth/status/1097889950865866752

[PrimeNumber7]

I don't want to discourage you from using the Lightning Network, but if you do it solely to earn on the routing fees expecting to make a decent income, I am afraid you will be disappointed. You won't be able to pay off the purchased hardware by routing payments. Someone correct me If I am wrong, but if you were to route hundreds of thousands of payments, you wouldn't even make $1. Because the money is routed through the cheapest channels. If your route is more expensive than the guy next to you, he will be the one whose channel is likely to be used.

This is true, owner of shitcoin.com perform such experiment with about 40BTC. In short, his node routed 389 transaction, but only earn about $0.34 during 2 weeks of experiment.

LN is still very new, and running at well below its capacity. As the number of economic transactions on LN grows, so will the fees that node operators can earn.

As the number of open channels a node operator has grows, so will the amount of fees it earns. LN payments are routed through the cheaper routes, but if you have more open channels with a diverse set of nodes, there will be a greater chance your channels will be among the cheapest route for transactions.

[vapapgr]

@Alex Bosworth report on  Feb 19, 2019 that his node routing more than $10,000/month and getting paid 0.25% for routing
if you buy hardware for $150 and routing more than $10,000/month you can have ROI after 4 months (lucky case)

https://twitter.com/alexbosworth/status/1097889950865866752

0.25% = $25 on $10k
you better not have more than 4 channels or else closing 4 channels each month to reaggregate will cost you 4x$3 to open and 4x$3 to close ($24)

so good luck operating a route with only 4 channels and breaking even

hey thanks for your time. Why do i have to close and re open the channels every month?

[mocacinno]

--snip--
hey thanks for your time. Why do i have to close and re open the channels every month?

You don't.
This is the funding tx of a channel on my node:
https://blockstream.info/tx/7caebcabf7757c85fd15f92c3ce4c6e5c1c79d6407cc2b966831f03609f700f0

It has been open since the end of 2018....

This one: https://blockstream.info/tx/71a3effc83b1b5664afb6103649b8c6f98c572ffe5f1a8a8abd6727d937cb40b
Also 2018...

This one: https://blockstream.info/tx/a435de3511e404c59781a84b9237c376e63c337fc0c3bca23b8711f7229e7556
Beginning 2019...

So, there is no technical reason why channels HAVE to be closed and re-opened every month... Offcourse, if i'd open a channel with somebody else, and i use this channel to keep on paying this certain individual (for example, for hosting), i'd have to close and re-open the channel once there's not enough funds left on my side... If i continued to pay said individual, my payment would no longer be routed trough the direct channel and i'd have to pay (minimal) routing fees AND have other open channels with funds on my side of the channel, unless some funds would move back to me (which could only happen if i got payed and the payment came from or was routed trough the person i routinely pay). But theoretical examples past this point get rather complex

[Pmalek]

<Snip>

I assume many of the current problems and those you described will be solved once the splicing feature gets implemented. It will allow users to deposit more money into an existing payment channel without having to make transactions on the first-layer network. Does anyone know if there is an ETA already for splicing?     

[vapapgr]

@Alex Bosworth report on  Feb 19, 2019 that his node routing more than $10,000/month and getting paid 0.25% for routing
if you buy hardware for $150 and routing more than $10,000/month you can have ROI after 4 months (lucky case)

https://twitter.com/alexbosworth/status/1097889950865866752

0.25% = $25 on $10k
you better not have more than 4 channels or else closing 4 channels each month to reaggregate will cost you 4x$3 to open and 4x$3 to close ($24)

so good luck operating a route with only 4 channels and breaking even

hey thanks for your time. Why do i have to close and re open the channels every month?

if you only have $10k invested. and you do $10k a month.. your funds are no longer yours in a specific channel path at the end of the month
before ever routing:
CustA($2500:$0)\
CustB($2500:$0)\\_you ($10000:$0) service
CustC($2500:$0)//
CustD($2500:$0)/

after doing $10k trade
CustA($0:$2500)\
CustB($0:$2500)\\_you ($0:$10000) service
CustC($0:$2500)//
CustD($0:$2500)/

as you can see. you now have $0 in you<>service on your side
yet... yes you have 4x$2500 on your side in the 4 custABCD<>you channels
but this cant help you be a payment path to a service .. so you need to reaggregate

if you want to continue being a path between customers and a service you have to close the channels and reaggregate the value from cust4x$2500. to $10k in the you<>service channel. and re assume new channels with new customers with funds.

you can stay open longer by having say $30k in the service channel. but then you would need:
either
   12 customers with $2500 to use up your $30k over 3 months
   but even then..
   12 channels closing after 3 months = $3onchain x 12 close + $3onchain x 12 open
   $72 costs.. and 30k over 3 months is still about $75 income
or
  4 customers with $7500 to use up your $30k in 3 months
  where then...
  4x$3onchain close 4x$3 onchain open
  which is $24 cost and $75 income

but its the balancing act of limiting channels and increasing funds of you. increasing funds customers have with you. to try to keep you active long enough to profit

so keep that in mind

Got it.
Still 1 question. When i open channel in Phoenix wallet i use 0.30 $ for 500 $ inbound/outbound.  Is it so expensive like15$ to open a channel on your own LN node?

[mocacinno]

--snip--

Got it.
Still 1 question. When i open channel in Phoenix wallet i use 0.30 $ for 500 $ inbound/outbound.  Is it so expensive like15$ to open a channel on your own LN node?

Opening and closing channels happens trough an on-chain transaction. At this moment, you'll have to pay ~40 sat/vbyte to get into the next couple of blocks... But the fee fluctuates quite a lot... So, yeah, at this moment in time a $15 fee would be possible...

[DaveF]

--snip--

Got it.
Still 1 question. When i open channel in Phoenix wallet i use 0.30 $ for 500 $ inbound/outbound.  Is it so expensive like15$ to open a channel on your own LN node?

Opening and closing channels happens trough an on-chain transaction. At this moment, you'll have to pay ~40 sat/vbyte to get into the next couple of blocks... But the fee fluctuates quite a lot... So, yeah, at this moment in time a $15 fee would be possible...

Side note when it comes to fees. IF you have funds in the on-chain wallet that is attached to your LN node then that is one thing.

IF you are following good security and don't keep funds in it then it's 2 transactions. Once to the wallet and the 2nd to open the channel. I do not keep funds in my on chain wallet that is on my node. So if I want to open a channel I 1st have to move some BTC to it.

Although I have had no issues with my node it's still somewhat beta so there is the possibility that something could go wrong.

I do have a backup of the wallet.dat file that my node uses so if it really dies I can get any on-chain funds off, and regularly backup my channels but the point is still there. Do you want to spend more time then needed recovering funds that could have been safer elsewhere.

-Dave

 

[vapapgr]

--snip--

Got it.
Still 1 question. When i open channel in Phoenix wallet i use 0.30 $ for 500 $ inbound/outbound.  Is it so expensive like15$ to open a channel on your own LN node?

Opening and closing channels happens trough an on-chain transaction. At this moment, you'll have to pay ~40 sat/vbyte to get into the next couple of blocks... But the fee fluctuates quite a lot... So, yeah, at this moment in time a $15 fee would be possible...

So for me that i ll open a channel for payments (i dont need it right now) i can do it with 12sat , right?

[mocacinno]

--snip--
So for me that i ll open a channel for payments (i dont need it right now) i can do it with 12sat , right?

Seems that at this moment, the mempool is not as full anymore... You could even get by with 7 or 8 sat/vbyte i guess.
The main thing to remember: opening and closing channels happens trough an on-chain transaction. If there's a fee-bidding-war going on because the mempool is overflowing, you'll pay a lot to open and close a channel. If the mempool is (allmost) empty, you might get by with a 1/2 sat/vbyte tx. This has nothing to do with the lightning network tough, it's because of the current size of the mempool (and the fees of the transactions in said mempool).

Once you have an open channel, you should be able to send funds allmost instantly with a very low fee... Even if they are routed trough other nodes. It seems my node usually ends up charging ~1,2 sat to route a payment...
Even if your payment gets routed over 10 nodes such as mine, you'll still pay between 10 and 20 sats, which is nothing compared to an on-chain transaction.


Wether it is worth routing payments is something completely different... I don't think i'll ever break even, but that was not my intention

[vapapgr]

--snip--
So for me that i ll open a channel for payments (i dont need it right now) i can do it with 12sat , right?

Seems that at this moment, the mempool is not as full anymore... You could even get by with 7 or 8 sat/vbyte i guess.
The main thing to remember: opening and closing channels happens trough an on-chain transaction. If there's a fee-bidding-war going on because the mempool is overflowing, you'll pay a lot to open and close a channel. If the mempool is (allmost) empty, you might get by with a 1/2 sat/vbyte tx. This has nothing to do with the lightning network tough, it's because of the current size of the mempool (and the fees of the transactions in said mempool).

Once you have an open channel, you should be able to send funds allmost instantly with a very low fee... Even if they are routed trough other nodes. It seems my node usually ends up charging ~1,2 sat to route a payment...
Even if your payment gets routed over 10 nodes such as mine, you'll still pay between 10 and 20 sats, which is nothing compared to an on-chain transaction.


Wether it is worth routing payments is something completely different... I don't think i'll ever break even, but that was not my intention

So helpful! Thanks.  My raspiblitz is on the way .... so i hope i ll contribute to that !

[zeuner]

I don't want to discourage you from using the Lightning Network, but if you do it solely to earn on the routing fees expecting to make a decent income, I am afraid you will be disappointed. You won't be able to pay off the purchased hardware by routing payments. Someone correct me If I am wrong, but if you were to route hundreds of thousands of payments, you wouldn't even make $1. Because the money is routed through the cheapest channels. If your route is more expensive than the guy next to you, he will be the one whose channel is likely to be used.

On the other hand, if you want to experiment with off-chain transactions and taking the load off from the main Bitcoin blockchain, that's a great thing to do.

Well, it's like experimenting with on-chain Bitcoin transactions in 2009. No one thought it could be for profit, and still it was technologically interesting enough to keep it up until it became profitable...