fix(rust_brain): preserve HLC on snapshot restore and gossip receive#42
Draft
cursor[bot] wants to merge 1 commit into
Draft
fix(rust_brain): preserve HLC on snapshot restore and gossip receive#42cursor[bot] wants to merge 1 commit into
cursor[bot] wants to merge 1 commit into
Conversation
restore_from_file() was rebuilding MemoryNode without the stored hlc field, assigning a fresh local HLC instead. After disaster recovery, causal writes from the original cluster (gossip replay, bulk_write) were rejected with TimestampRegression — silent data loss in production failover paths. Gossip receive() also omitted hlc, allowing stale remote events to overwrite fresher local state when monotonic enforcement compared against a fresh HLC. - Add _parse_hlc() for snapshot/wire normalisation with legacy fallback - Restore hlc and advance global _hlc during snapshot restore - Pass hlc through bulk_write; lock restore and gc_expired - Gossip: propagate hlc, reject stale updates, skip missing-hlc overwrites - Tests for restore round-trip and gossip HLC semantics Co-authored-by: Daniel <DJLougen@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Bug and impact
Critical data-loss bug in disaster recovery and multi-node sync paths.
restore_from_file()rebuiltMemoryNodewithout the storedhlcfield, assigning a fresh local HLC. After snapshot restore, causal writes from the original cluster (gossip replay, bulk replication) were rejected withTimestampRegression— silent data loss during failover.GossipProtocol.receive()also omittedhlc, allowing stale remote events to overwrite fresher local state.Concrete trigger
remember("session:1", data, hlc=(5000, 10, "nodeA"))hlc=(5000, 11, "nodeA")→ rejected (restored node has fresh HLC)Root cause
v0.6.0 added HLC monotonic enforcement but
restore_from_file()andgossip.receive()were not updated to preserve/apply HLC timestamps.Fix
_parse_hlc()for snapshot/wire normalisation with legacy fallbackhlcand advance global_hlcduring snapshot restorehlcthroughbulk_write; lockrestoreandgc_expiredhlc, reject stale updates, skip missing-hlc overwrites on existing keysValidation
test_restore_preserves_hlc_for_replication— HLC round-trip + causal successor writetest_enterprise_backup,test_gossip,test_rust_brain)