[RIP-81] Master-slave metadata incremental synchronization - #9617
Merged
Conversation
2. Increased snapshot storage interval to reduce the increased garbage collection frequency caused by deep copies 3. Added DataVersion transmission
RongtongJin
approved these changes
Oct 21, 2025
RongtongJin
merged commit Oct 21, 2025
e20354e
into
apache:feature-sync-metadata
11 of 12 checks passed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Which Issue(s) This PR Fixes
Fixes #9616
Brief Description
Master Node:
MetadataChangeObserver serves as a monitoring interface, extending two subclasses, NoopMetadataChangeObserver and SyncMetadataChangeObserver, for implementing no-op and synchronous message sending, respectively. Configuration map operations in configuration classes such as TimerMetrics, MessageRequestModeManager, ScheduleMessageService, ConsumerOffsetManager, SubscriptionGroupManager, and TopicConfigManager have been modified: Compute methods are used to ensure the atomicity of message sending and in-memory configuration updates, ensuring simultaneous success and failure to prevent data inconsistencies.
For the frequently updated consumerOffset, a dual buffering approach is implemented, with scheduled and buffer-full synchronous message sending strategies. This reduces the synchronous message sending burden caused by frequent consumerOffset updates. The scheduled strategy also mitigates the long inconsistency window caused by the buffer-full strategy in low-frequency updates.
A scheduled task is used to store the latest metadata snapshot, which contains the current metadata and the maximum offset of the corresponding system topic.
Methods for obtaining snapshots from the master node are defined in the AdminBrokerProcessor and BrokerOuterAPIs.
Slave nodes:
How Did You Test This Change?
broker config
allowMetadataIncrementalSync = trueRelevant work is already in progress. Proposal documentation is available here:
English version:
https://docs.google.com/document/d/1Q4dckYs5ALt_BM5aId9-3N9-kaGL-cLYvAVhv9Mc7bs/edit?usp=sharing
Chinese version:
https://m0sgzrp00oj.feishu.cn/docx/CvbedEJXto7A2BxfYVXcLXATnbd?from=from_copylink