fix(cluster): recover writes after a fast partition primary rejoin - #3987
Merged
Conversation
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## master #3987 +/- ##
============================================
- Coverage 84.63% 83.69% -0.94%
- Complexity 1398 1405 +7
============================================
Files 1224 1224
Lines 176637 177971 +1334
Branches 143161 144285 +1124
============================================
- Hits 149491 148956 -535
- Misses 23173 25076 +1903
+ Partials 3973 3939 -34
🚀 New features to boost your workflow:
|
numinnex
reviewed
Aug 28, 2026
numinnex
reviewed
Aug 28, 2026
numinnex
reviewed
Aug 28, 2026
numinnex
reviewed
Aug 28, 2026
numinnex
reviewed
Aug 28, 2026
numinnex
reviewed
Aug 28, 2026
Contributor
Author
|
/ready |
numinnex
reviewed
Aug 28, 2026
numinnex
reviewed
Aug 28, 2026
numinnex
reviewed
Aug 28, 2026
Contributor
Author
|
/ready |
numinnex
approved these changes
Aug 28, 2026
hubcio
approved these changes
Aug 28, 2026
hubcio
added a commit
to chengxilo/iggy
that referenced
this pull request
Aug 31, 2026
Merging master pulled in the partition forward path (apache#3987), written against the pre-branch roster API: replica_ip() as an Option, ClusterRoster.self_ip, and an infallible From for ResolvedClusterNode. This branch validates roster IPs at config time, so replica_ip() is plain and the conversion is a TryFrom. Git merged the hunks without a textual conflict and left the server crate uncompilable. Adapt the forward path and its test helper to the resolved roster. The unparsable-IP case in the roster walk test is gone: such a node can no longer exist in a resolved roster.
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
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.
A fast partition primary rejoin could leave the group unable to
commit new writes. A backup could adopt suffix headers without
obtaining their bodies because repair stopped at commit_max. Clients
then handled the typed non-admission response by checking metadata
leadership, even though metadata and partition groups can elect
different primaries.
Extend suffix repair to the available frontier and clear completed
sessions after the later commit catches up. Reject writes on stale
backups before local offset validation, so callers receive the typed
response required for safe rerouting.
Add one-pass roster routing for HTTP and the Rust, Node, Go, Java,
and C# clients. Coordinate concurrent Rust reconnects and prevent
ambiguous replicated writes from being replayed under a new session.