Keyspace Partitioning for Dynamic Storage Capacity Management

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing distributed databases like Cassandra face challenges in efficiently managing storage capacity, as the recommended storage capacity per node can be exceeded, requiring frequent additions or replacements of resources.

Innovation Solution

A system and method for partitioning a keyspace that includes a server transmitting a partition setting item based on a keyspace creation request, which includes a replication factor and parameters such as the size of a partition for data storage or the storage period of data. This allows for dynamic management of storage capacity and resource allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If the storage capacity per node is increased to handle more data, then the data storage capability is improved, but the cost of adding or replacing hardware resources increases

Engineering Contradiction:

Improvedata storage capacityVSAvoidhardware resource management

Core Design Contradiction:

Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent divides the storage system into multiple partitions within a keyspace, allowing data to be distributed across different nodes. Instead of increasing storage capacity on a single node, the system segments data storage across multiple nodes, each maintaining manageable storage levels while collectively providing expanded storage capacity.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a new dimension of storage management by implementing partition-level control with configurable retention periods and replication factors. This moves from single-node capacity management to multi-dimensional control across nodes, partitions, and time periods, allowing storage expansion without proportionally increasing hardware complexity.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Reliability

If the replication factor is increased to improve data reliability, then the system reliability is improved, but the storage capacity consumption increases

Engineering Contradiction:

Improvedata reliabilityVSAvoidstorage capacity

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent makes the replication factor dynamic rather than fixed. Different partitions can have different replication factors based on their specific reliability requirements and data importance. This allows the system to optimize the balance between reliability and storage capacity consumption on a per-partition basis, rather than applying a uniform replication factor across all data.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent applies different replication factors to different partitions based on their specific needs. Critical data partitions can have higher replication factors for enhanced reliability, while less critical partitions can use lower replication factors to conserve storage capacity. This local differentiation resolves the contradiction by tailoring reliability measures to actual requirements.

Inventive Principle:

Principle #3Local quality

3Duration of action of stationary object

If the data retention period is extended to preserve historical data, then the data preservation capability is improved, but the storage capacity requirement increases

Engineering Contradiction:

Improvedata retention periodVSAvoidstorage capacity

Core Design Contradiction:

Duration of action of stationary objectVSQuantity of substance

Solution Approach 1:

The patent segments data retention by creating different partitions with different retention periods. Instead of retaining all data indefinitely across the entire system, hot data can be retained in partitions with shorter retention periods while cold historical data is moved to partitions with longer retention periods. This segmentation allows the system to preserve historical data while managing storage capacity through differentiated retention strategies.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent implements periodic data lifecycle management where data automatically transitions between partitions based on retention period expiration. This periodic action allows the system to preserve historical data in archived partitions while automatically freeing up storage capacity in active partitions, resolving the contradiction between long-term preservation and storage capacity management.

Inventive Principle:

Principle #19Periodic action

Data Source

PatentUS12326848B2System, method and computer readable storage medium for partitioning keyspace

Publication Date: 2025.06.10 SAMSUNG SDS CO LTD

AI summary

A system for partitioning a keyspace includes: a system for partitioning a keyspace, includes: a server configured to transmit a partition setting item based on a keyspace creation request; and a database configured to receive the partition setting item and configured to create a first partition for the keyspace based on the received partition setting item, wherein the partition setting item includes a replication factor and at least one of a size of a second partition for storing data or a storage period of the data.