LND adds 10,000 sat/vB cap to external fee estimates
LND’s October 22 release candidate focuses on limiting risk from external fee estimates and improving node reliability. The build sets an upper bound for remote fee inputs and fixes a native SQL connection-pool deadlock. It is a pre-release that requires no database migration and includes verification instructions, which differs from real-world deployment testing.
TL;DR
- Version v0.21.5-beta.rc1 is a pre-release.
- A ceiling now limits extreme external fee estimates, and the minimum-relay estimate is bounded.
- Verification confirms downloads, while deployment testing validates behavior in actual node setups.
External fee inputs now have strict bounds
The release notes document that external estimates previously had only a lower bound. The change adds an upper bound before the value reaches fee-sensitive wallet operations.
The change sets a ceiling of 10,000 satoshis per virtual byte and also bounds the API’s minimum-relay estimate. The ceiling acts as a guardrail against unusable values caused by errors, data format changes, or compromised dependencies. Fee estimates serve as inputs to decisions, not instructions, and bounding them limits the impact of upstream issues on a wallet.
The broader operational theme aligns with Bitcoin Core’s revised disclosure process: financial software needs both systematic problem discovery and a path to deploy fixes safely.
Pre-release status and testing requirements
The release page marks the build as a pre-release and states no database migration is required. It provides signatures and binary verification steps that confirm artifact integrity. Those steps do not demonstrate behavior across diverse node configurations.
The notes include a native SQL connection-pool deadlock fix. Reliability depends on storage behavior as much as network communication. A process that stays up while requests stop completing still fails dependent applications.
Testing must reflect real operations: receiving requests, accessing storage, handling failures, and restarting cleanly. The lack of a database migration removes one category of change, yet it does not prove that all interactions with existing data remain unchanged.
Operational reliability is separate from custody design. This distinction applies to topics like shared-key Bitcoin lending and custody and recovery arrangements. Secure signing authority does not replace ongoing software maintenance. This candidate introduces specific changes to evaluate. The final release decision and each operator’s deployment decision remain separate steps.
Authors credited by the source: NewsBTC Editorial Team
Published by Holder based on an external source.



