[server] Register external KV snapshots and restore replicas - #4242
Open
platinumhamburg wants to merge 5 commits into
Open
[server] Register external KV snapshots and restore replicas#4242platinumhamburg wants to merge 5 commits into
platinumhamburg wants to merge 5 commits into
Conversation
Share KV snapshot file metadata across modules and register externally produced target-bucket snapshots with coordinator fencing and idempotent retries. Adopt confirmed snapshots into retention and protect persistent references when registration responses are lost. Initialize the local log boundary from snapshot-only history through ordinary replica role notifications. Cover external snapshot production, registration, recovery, repeated notifications, online writes, failover and subsequent snapshot retention.
Use the selected KV snapshot to initialize an empty local log after its files have been loaded, and capture the leader log boundary after recovery. Reject conflicting log initialization and propagate snapshot lookup errors. Release partially recovered KV state before retrying, and publish leader identity only after recovery succeeds so failed elections can be retried. Cover snapshot lookup and download failures, metadata retry after loading, and protection of existing local log records.
Keep the existing leader and follower transition behavior while coordinating KV snapshot recovery with local log initialization. Test snapshot lookup failures independently of role transition retries, and recover from a failed download under a subsequent leader epoch.
Serialize external snapshot identity and ordering checks, persistent handle confirmation and retention updates in CompletedSnapshotStore. Delegate coordinator fencing and idempotent persistence to the handle store while preserving ordinary snapshot writes and cleanup behavior. Describe the snapshot counter check as an allocation bound; reserving the ID remains the caller's responsibility.
Resolve the recovery snapshot and log prerequisites before changing role state or discarding the current KV. Reuse that selection for the first recovery attempt while preserving retries through older healthy snapshots. Avoid remote manifest lookups when the local log already covers the snapshot. Cover metadata lookup failures with a live KV and recovery after metadata becomes available. Clarify that obsolete snapshot cleanup may be deferred.
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.
Purpose
Linked issue: close #4223
Brief change log
Tests
API and Format
Documentation