Skip to content

User Story 2026_051

As a Node Operator, I want replays and replay requests between a CASE_MANAGER and a peer that is out of step to be rate-bounded, so that a peer that stays out of sync cannot flood either node

Notes

A peer that keeps rejecting entries, or keeps asking for a case it cannot seed, would otherwise trigger a replay for every message. Bounding the rate, backing off on repeats from the same position and resetting the bound when the position moves keeps recovery working while it caps the traffic. Each limit lets the next attempt through after a bounded interval, so recovery is never blocked for good.

Metadata

  • Speaker: Implementer (Node Operator)
  • Related issues: #4487 (source Concern: the user-story set is stale), #2599 (source Idea: distributed case consistency and replication)
  • Drafted for: #4516 (split of the catch-all story 2022_088)
  • Added: 2026-11-03