Graph Synchronizer
The GraphSynchronizer serves as a core component of the graph management system, bridging the gap between transient
in-memory graph structures and their corresponding persistent storage representations. Its central function revolves
around ensuring each node in a provided graph is correctly reflected in the storage system.
Graph Synchronization
The process begins when a graph is handed over to the GraphSynchronizer. For each node in this graph,
GraphSynchronizer leverages the NodeIdentificatorsProvider to ascertain the existence of a corresponding
node within the persistent storage. These identifiers are customizable and specific to each node type, giving
you the flexibility to define them as required.
Once a node's existence is determined, the GraphSynchronizer takes one of two actions:
-
Node Merging: If the node already exists in the storage, the new node and the existing one are merged. During this merging process, several default operations are carried out:
-
Attribute Merging: If the new node introduces attributes that are not present in the existing node, these are added to the stored node, ensuring it is up-to-date.
-
Node De-duplication: If multiple nodes in the new graph share the same identifier, these are consolidated into a single node in the storage, preventing redundancy and promoting efficiency.
-
Edge Merging: In the case of duplicate nodes with edges, the edges are also merged appropriately during the node merging process.
-
-
Node Creation: If the node does not exist in the storage, the new node is added to the persistent storage as is.