Distributed Database Metadata Sync via Full Replica Acknowledgment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In large-scale distributed databases, synchronously updating metadata across database nodes is challenging, leading to inconsistencies where some nodes read updated metadata while others read outdated metadata.
Innovation Solution
A method and apparatus utilizing a full synchronization protocol to synchronize metadata updates through a primary replica and backup replicas, employing update markers such as version numbers or hashed metadata to ensure all nodes read the latest metadata.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If metadata update data is pushed to database nodes in a distributed database, then the metadata can be updated across nodes, but it cannot be ensured that update data is simultaneously pushed to all nodes, causing inconsistency where different nodes read different metadata
Solution Approach 1:
The patent applies preliminary action by requiring the primary node to receive acknowledgment confirmations from all backup nodes before completing the metadata update. This ensures that the update is propagated to all nodes before the system considers the update complete, preventing inconsistency. The primary node waits for all backup nodes to confirm receipt of the update data, thereby ensuring simultaneous availability across the distributed system.
2Reliability
If a full synchronization protocol is used between primary replica and backup replicas, then all nodes can read updated metadata, but the complexity of the update process increases
Solution Approach 1:
The patent implements feedback through a confirmation mechanism where backup nodes send acknowledgment messages to the primary node after receiving update data. The primary node uses this feedback to determine when all backup nodes have successfully received the update. This feedback loop ensures reliability by verifying that the synchronization is complete before the update is considered successful, while keeping the protocol manageable through structured acknowledgment exchange.
Data Source
AI summary
Embodiments of this specification provide a method and an apparatus for synchronously updating metadata in a distributed database. The distributed database includes a plurality of database nodes. A first node stores a first primary replica of a target replicated table, and a plurality of remaining second nodes store first backup replicas of the target replicated table. A full synchronization protocol is used between the first primary replica and the first backup replicas. The method is applied to the first node and includes: receiving an update request related to first metadata of a first data table in the distributed database; and writing update information related to the first metadata into the first primary replica based on the update request, to synchronize the update information to the respective first backup replicas of the plurality of second nodes based on the full synchronization protocol.