Distributed Graph Database Backup via Shard Export and Intermediate Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data processing methods for distributed graph databases lack flexibility, practicality, and efficiency in data backup and restoration processes.
Innovation Solution
A data processing method and apparatus that involves storing graph data in shards within a first cluster, determining target storage nodes based on graph topology structure information, exporting these shards to an intermediate storage device, and storing them in a second cluster based on independent topology structure information, decoupling the data export and import processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data backup is performed using existing methods for distributed graph databases, then data security and reliability are ensured, but the process lacks flexibility and efficiency
Solution Approach 1:
The patent segments graph data into shards and distributes them across multiple storage nodes in different clusters. Each shard can be independently backed up, transferred, and restored, enabling flexible and efficient data backup operations without requiring complete system downtime or complex coordinated transactions across the entire database.
2Productivity
If data is transferred directly from source to destination without intermediate storage, then the process is simpler, but cross-area network overheads increase and reduce efficiency
Solution Approach 1:
The patent introduces an intermediate storage device as a mediator in the data transfer process. Data is first exported from source storage nodes to this intermediate device, then imported to destination storage nodes. This intermediary approach optimizes network utilization, reduces cross-area network overheads, and enables parallel transfer operations that improve overall data processing efficiency.
3Adaptability or versatility
If cluster topology structures are coupled during data import, then the process is more constrained, but flexibility in choosing target cluster configurations is reduced
Solution Approach 1:
The patent extracts the data from its original cluster context and stores it in an intermediate device in a cluster-independent format. This allows the data to be imported into destination clusters with different topology structures without requiring the source and destination clusters to have matching configurations. The extraction of cluster-specific dependencies enables flexible target cluster configuration selection while maintaining straightforward import operations.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
Embodiments of this specification provide a data processing method and apparatus for a distributed graph database. In the data processing method for a distributed graph database, graph data of the distributed graph database are stored in a form of a shard in a storage node in a first cluster. The data processing method includes: determining to-be-backed-up graph data; determining at least one target storage node based on a correspondence between a graph data shard and a storage node, in the first cluster, in which the graph data shard is located; exporting several graph data shards in the target storage node to an intermediate storage device; and then storing the graph data exported to the intermediate storage device in at least one storage node in a second cluster based on a correspondence between the graph data shard and a storage node, in the second cluster, in which the graph data shard is to be stored.