Lightweight Graph Database with Separate Searchable Datastore
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Graph databases face scalability issues and increased search latency due to the addition of properties to nodes, which can lead to exponential growth in database size and traversal time.
Innovation Solution
A computer-implemented method that creates a lightweight graph database and a separate searchable datastore, where the payload of messages is stored in the datastore and only essential information is stored in the graph database, preventing excessive growth and maintaining fast traversal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If properties are added to entity nodes in the graph database, then the information storage capability is improved, but the database size increases exponentially and traversal speed decreases
Solution Approach 1:
The patent segments the data storage into two separate systems: a graph database for storing entity relationships and a separate searchable datastore for storing message payloads and properties. This segmentation prevents the graph database from growing exponentially while maintaining fast traversal capabilities, as the graph database only contains essential entity and relationship data without bulky property information.
Solution Approach 2:
The patent extracts message payloads and detailed properties from the graph database nodes and stores them in a separate searchable datastore. This extraction eliminates the exponential growth problem in the graph database while preserving full information availability through the separate datastore, allowing fast graph traversal without information loss.
2Loss of information
If message payloads are stored in the graph database, then data completeness is improved, but the database size increases and search latency increases
Solution Approach 1:
The patent segments data storage by placing message payloads in a separate searchable datastore rather than storing them in graph database nodes. This segmentation maintains data completeness through the separate datastore while preventing search latency in the graph database, as traversal operations only query the compact graph structure without scanning large payload contents.
Solution Approach 2:
The patent introduces a separate searchable datastore as an intermediary between the graph database and the message payloads. This intermediary stores complete payload information while allowing the graph database to remain lightweight and fast for relationship traversal, effectively mediating between data completeness requirements and performance requirements.
3Adaptability or versatility
If properties are represented as nodes in the graph, then property querying capability is improved, but the graph database size increases exponentially
Solution Approach 1:
The patent segments property storage by moving property data to a separate searchable datastore while keeping the graph database focused on entity relationships. This segmentation maintains property querying capability through the separate datastore without causing exponential growth in the graph database size, as properties are no longer represented as additional nodes in the graph structure.
Data Source
AI summary
A computer-implemented method includes receiving a message comprising an origin, a destination and a relationship type for a relationship between the origin and the destination. The message further includes a payload. A first node is created in a graph database for the origin and a second node is created in the graph database for the destination. A relationship is set between the first node and the second node in the graph database based on the relationship type. A node is created in the graph database for the message while preventing the payload from being stored in the graph database. A relationship is created between the first node and the node for the message. The message, including the payload, is stored in a searchable datastore separate from the graph database.