Navigating the lightning network with Explorer
Data-driven lightning network explorer by bolt.observer
Information wants to be free. Or at least browsable. In this spirit, we recently released our Explorer which enables you to drill down into lightning networks’ graph data, explore nodes close and afar, find the next node you want to open a channel to or just learn something interesting about the lightning network and its participants.
We at bolt.observer like data. What we like even more is using data to help us make better decisions, improve our processes and teach us something new or interesting.
Deploying new bitcoin into the lightning network can mean a daunting decision - depending on your goals your optimal channel peers can be significantly different. The steady growth of the lightning network means that we have more and more choices, and picking the right node for your precious bitcoin is no easy feat with close to 20k nodes in the network. With Explorer you are able to filter nodes in the network by many different parameters - from basic ones like the number of channels and total capacity to slightly more advanced ones like median and average channel sizes, fee structures, or connectivity types.
But we didn’t want to stop here. We wanted to put our users (well, their nodes) in the center of attention. Using our reference node filter you can explore the network from your (nodes) point of view, enabling node operators to observe how nodes close to them look and behave.
We developed Explorer partially to solve our internal needs for deployment of our monitoring infrastructure, but mostly because we love exploring lightning data (as you might have guessed from our #lightningfact twitter series).
Explorer
Explorer gives you the ability to explore lightning network data with various filters and thus enabling you to think about it more analytically. Besides standard parameters provided through gossip (channels, capacity, fees, etc) you can also use many of our aggregates (average/median/maximum) or data that we obtain by crawling the network ourselves - for example, channel size limits. More of those will be announced in the following weeks.
When visiting https://bolt.observer/explorer you are presented with this advanced search form:
A list of all available parameters that can be used in filters can be found in our documentation. For each of the parameters you have three available conditions with which you can narrow down your searches precisely:
Less than
Equals
Greater than
The only exception to this is supported communication options - ipv4, ipv6, and tor.
Let’s explore some ideas on how we can use various filters and what information can be learned from their results.
Hops
With our reference node search, you can easily see what kind of nodes you have in your neighborhood, their inbound and outbound fees and capacities. This can give you better insight into what is happening around you, and how similar or different you are from your neighbors. It also enables you to see how much of the network you can reach within a certain number of hops - showing you how well connected you are. A lower number of hops to reach the majority of the network means that generally, your payments could/will be faster and more reliable (more hops = more potential problems).
In our previous article, we talked about the concept of the neighborhood. To quickly remind you what we meant by that - neighborhood consists of nodes you have channels with (first hop), and nodes that they are connected to (second hop). If you feel very adventurous you can also include nodes that are 3 hops away from you but by then, if you have more than a couple of channels, you will already cover the majority of current nodes.
Connectivity
As I’ve already mentioned in our previous article - how a node communicates with the rest of the network, mainly tor or clearnet makes a difference. While privacy gains by using tor are significant downsides are higher latency and lower reliability of the network. This tradeoff means that depending on your use case and threat model you might want to prioritize opening channels to nodes that are using clearnet connections (either ipv4 or ipv6) over nodes that are using tor. Without Explorer, you can easily filter out nodes by their advertised address types.
Capacity
Sending large(er) payments requires channels with sufficient capacity over the entire path (not accounting for MPPs) which means that opening channels to nodes that have sufficiently large channels will make your life easier. In Explorer we offer filters for three channel parameters:
Average channel size
Median channel size
Maximum channel size
Let's take a look at River Financial node as an example:
By comparing all three numbers you can gain insight into how their channel size distribution looks and make more informed decisions about where to deploy your precious bitcoin.
An additional useful data point is the minimum channel size requirement by a specific node. In River’s example, this is set to 1 million sats which means that generally, you can expect all channels to be at least of this size (depending on when the limit was configured smaller channels can exist).
Fees
Last but not least is fees. We all want to earn them, but we prefer not to pay (too much) of them. With Explorer, you can find potential channel peers that fit your needs, for example finding nodes that support zero base fee movement. Same as with channel capacity you also have the ability to filter outgoing or incoming fees by three parameters:
Average fee rate/base fee
Median fee rate/base fee
Maximum fee rate/base fee (outgoing only)
We’ve seen several examples of how we can use network data and filtering to find nodes that match our desired criteria, but there is another side to Explorer that I’ve mentioned in the beginning. If you are anything like me, you are a curious individual with a deep interest in the lightning network, then you might also enjoy Explorer just for the sheer ability to explore this great invention we call lightning network from a data perspective. For example - did you know that about a third of the network is just single channel nodes? I invite you to find something interesting yourself and share it with us over Twitter!