Highly Available Lightning Channels Revisited – ROUTE OR OUT

In a detailed discussion with a routing node operator, several insights and strategies for managing high-availability (HA) nodes within a network were explored. The conversation highlighted that it might not be essential to signal high availability per channel, suggesting that indicating this status at the node level in the announcement might be sufficient. An alternative method, without requiring code changes, could involve using an alias suffix such as (HA) to denote high availability.

The dialogue also touched upon the flexibility regarding the signaling of a guaranteed amount for routing. It was mentioned that rather than specifying a fixed amount, the responsibility could be left to node operators to ensure they have adequate liquidity to handle the traffic they encounter. This approach allows for a more dynamic management of resources based on actual needs. Additionally, as part of managing a HA node, there could be a strategy to stop signaling channels that have been depleted to optimize the network's efficiency.

Operators are beginning to experiment with methods to avoid penalization for failed transactions. Instead of returning the typical "temporary channel failure" error, some are exploring the use of different errors or even corrupt failure messages that cannot be attributed to evade penalization more effectively. This experimentation indicates a shift towards more sophisticated handling of errors to maintain node efficiency and reputation.

Another aspect covered was the temporary cessation of routing by some nodes, particularly during periods of high chain fees, to avoid the risk of forced closures. This strategy underscores the financial considerations node operators must balance to maintain service while managing costs.

The conversation also revealed a preference for selecting HA nodes for blip39 blinded paths, aiming to increase the likelihood of successful transactions. This preference suggests a recognition of the importance of reliability in node selection criteria.

Finally, the discussion acknowledged the importance of proper inbound liquidity as a criterion for being considered a HA node. This requirement extends the responsibility of the node operator to ensure they can consistently meet the demands placed on their node. However, challenges remain in identifying performance issues, especially with nodes operating over Tor, which are noted to be significantly slower without a current solution to pinpoint these nodes accurately.

These insights reflect ongoing efforts to refine the management and operation of nodes within networks, focusing on reliability, efficiency, and strategic liquidity management to enhance overall network performance.