Rebalancing argument [combined summary]



Individual post summaries: Click here to read the original discussion on the lightning-dev mailing list

Published on: 2018-07-11T07:56:20+00:00


Summary:

The Lightning Network community has been engaged in discussions regarding the need for active rebalancing in the network. The importance of maintaining channel balances to prevent routing failures and nodes fighting over fees was highlighted. One participant argued that rebalancing can be avoided if nodes plan their send and receive activities based on available balances. Another perspective emphasized the benefits of distributing available balances equally among a node's channels to minimize depleted channels and reduce routing failures.While passive rebalancing may not be necessary in a post-AMP (Atomic Multipath Payments) world, it may still be required if certain sources/sinks dominate, resulting in biased channels that require maintenance. The directionality of payment flows also plays a role in determining the necessity of active rebalancing. If payment flows are balanced, then passive rebalancing may not be needed. However, if a few sources/sinks dominate, regular rebalancing might be necessary.A recent paper suggests that the distribution of funds in channels does not significantly impact network liquidity, challenging the notion of rebalancing. Instead, the author suggests that a smart fee strategy could achieve the desired balance by allowing nodes' plans to send and receive payments to dictate the distribution of funds. Nodes with no balance are considered detrimental to the network's topology and should be discouraged.In a discussion between Dmytro and Olaoluwa, Dmytro questions the necessity of equilibrium in the Lightning Network, stating that there is no concrete evidence that it improves the network. However, he acknowledges that locked funds in channels pose a major issue, and the distribution of funds should be influenced by nodes' plans rather than equilibrium. Olaoluwa argues that active rebalancing may be necessary for nodes to facilitate their own payments, while passive rebalancing can ensure an equilibrium state among available balances. The directionality of payment flows also influences the need for passive rebalancing. If payment flows are balanced, passive rebalancing might not be necessary. Dmytro suggests that a smart fee strategy could potentially replace the need for rebalancing.Researchers and developers have been exploring the topic of rebalancing in the Lightning Network. Some argue that sending cycle transactions to oneself may not be necessary as the distribution of funds in channels does not significantly impact network liquidity. Instead, a smart fee strategy is proposed as an alternative solution. The current fee schedule for c-lightning and lnd already factors in the impact of time locks on edge weight during path finding. Future developments might allow nodes to signal how they want fees to scale with the absolute CLTV (CheckLockTimeVerify) of HTLC extend. Additionally, introducing a "balance disruption" factor in the fee schedule could penalize unbalancing a channel, which would help quantify apprehension towards time locks and channel balance equilibrium affinity.The discussion on active rebalancing in the Lightning Network highlights its necessity when nodes require more capacity to receive payments. Pre-AMP, active rebalancing was crucial, but multi-path payments have made it less necessary. Passive rebalancing can help reduce routing failures from failed HTLC forwarding by maintaining an equilibrium state among channel balances. The directionality of payment flows also affects the need for passive rebalancing. If payment flows are balanced, passive rebalancing may not be essential. Existing literature shows that active rebalancing is not a new problem and has been explored previously.Dmytro Piatkivskyi questions the benefits of rebalancing the Lightning Network through cycle transactions, suggesting that a smart fee strategy could achieve cheaper fees and shorter paths without the need for rebalancing. He also argues that the distribution of funds in channels may not significantly impact network liquidity. However, Rene Pickhardt points out that the Lightning Network setting differs from a 2010 paper on transaction routing models, and routing fees invalidate certain theorems and lemmas. Pickhardt suggests Atomic Multipath Payments and splicing as potential solutions to the issues raised. The use of HTLCs affects network flow by locking up funds during routing. Privacy requirements and the growing size of the network may pose further challenges in its routing.An email exchange between Rene Pickhardt and Dmytro Piatkivskyi expands on the discussion surrounding rebalancing in the Lightning Network. Piatkivskyi suggests that a smart fee strategy might be more effective than sending cycle transactions for balancing the network. He references a paper stating that the distribution of funds in channels has minimal impact on network liquidity. However, Pickhardt argues that balanced channels increase the chances of finding a working route, mentioning Atomic Multipath Payments and splicing as potential solutions. Routing fees and HTLCs are highlighted as factors that complicate the problem of payment routing. Various links provided offer further reading on the topic.The author of a post questions the benefits of rebalancing the Lightning Network through cycle transactions and suggests that a smart fee strategy could achieve cheaper fees or shorter paths without rebalancing.


Updated on: 2023-07-31T20:23:07.437462+00:00