Client-Side Index for Distributed Database Search

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing database systems face limitations in network bandwidth, throughput, and computational performance, particularly when handling non-key-value queries in distributed databases, leading to increased latency and resource utilization issues.

Innovation Solution

The solution involves shifting query handling from the database server to client applications, where the index is moved from the database server to the network node clients, allowing for local search operations and reducing the load on the database server, thereby improving search efficiency and reducing latency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the index is maintained at the database server for non-key-value queries, then query capability is provided, but network bandwidth and server computational performance are limited

Engineering Contradiction:
Improvequery capabilityVSAvoidnetwork bandwidth utilization
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

The index data structure is extracted from the database server and placed at the client application. The client maintains a local copy of the index, allowing it to perform search operations independently without querying the server, thereby eliminating network bandwidth limitations while preserving full query capability.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces an index data structure as an intermediary between the client and the database server. This intermediary contains pre-computed search keys and their corresponding data post identifiers, enabling the client to perform local binary search operations without direct server involvement for search queries.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Ease of operation

If the index is maintained at the database server, then centralized management is achieved, but server computational load increases

Engineering Contradiction:
Improvecentralized managementVSAvoidserver computational load
Core Design Contradiction:
Ease of operationVSPower

Solution Approach 1:

The client application is empowered to maintain its own local index data structure. When data posts are added, modified, or deleted, the client autonomously updates its local index by receiving notifications from the server and performing the necessary index maintenance operations locally, eliminating the need for the server to handle index maintenance computational load.

Inventive Principle:
Principle #25Self-service

3Reliability

If distributed query is performed across all database instances, then complete search coverage is achieved, but query latency increases

Engineering Contradiction:
Improvesearch completenessVSAvoidquery latency
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The index data structure is pre-computed and maintained at the client application before search operations are needed. The index contains all necessary search keys and their corresponding data post identifiers, allowing the client to perform instant local binary search operations without needing to query multiple database instances, thus achieving both search completeness and low latency.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3649532B1Methods, systems, databases and network nodes of data communication networks for handling data posts
Publication Date: 2022.12.07 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)
  • EP3649532B1 patent drawingFigure 1
  • EP3649532B1 patent drawingFigure 2
  • EP3649532B1 patent drawingFigure 3~4

AI summary

A method is disclosed performed by a system (300) of a data communication network, the system comprising a database (310), a first and a second network node (321, 322), the database (310) operating as a server with respect to the first and the second network nodes (321, 322) operating as clients to the database (310), the system handling data posts comprising one or more data values of different categories and a unique ID of the data post. The method comprises sending, by the first network node (321) to the database (310), a request for adding a first data value and a first ID to the database (310), storing by the database (310) the first data value and the first ID, sending, by the database (310) to the second network node (322), a message comprising the first ID and the first data value and storing, by the second network node (322), the first ID linked to the first data value, thus making the data stored at the database (310) searchable at the second network node (322).