Virtual Node Storage Abstraction for Seamless Host Migration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In distributed computing environments, managing virtual nodes is challenging due to difficulties in allocating resources and migrating them between hosts, especially during device failures, software updates, or load balancing, which can lead to downtime and inefficiencies.

Innovation Solution

The solution involves deploying virtual nodes with both non-persistent and persistent storage components, where non-persistent data is stored locally on the host and persistent data is stored remotely, allowing for seamless migration and resource allocation across hosts while maintaining minimal downtime. A management system identifies transition events and selects new hosts based on load balancing and resource requirements, ensuring continuous operation by initiating execution on the new host using the same persistent data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If virtual nodes are migrated between hosts during device failure, software updates, or load balancing, then system reliability and resource utilization are improved, but migration downtime and operational disruption increase

Engineering Contradiction:
Improvesystem reliabilityVSAvoidmigration downtime
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent implements preliminary action by pre-configuring persistent storage for virtual nodes before migration events occur. The system establishes storage abstraction layers and pre-positions data in accessible locations, so when migration is triggered by device failure, software updates, or load balancing, the virtual node can be quickly instantiated on a new host without waiting for data transfer, thereby reducing migration downtime while maintaining system reliability

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary storage abstraction layer between the virtual node and physical storage resources. This intermediary layer decouples the virtual node from host-specific storage, enabling seamless migration by mediating data access across different hosts. The abstraction layer handles data location transparency, allowing virtual nodes to migrate between hosts without direct data movement, thus reducing migration downtime while ensuring system reliability

Inventive Principle:
Principle #24Intermediary (Mediator)

2Productivity

If multiple virtual nodes are deployed on a single host to increase resource efficiency, then resource utilization improves, but resource allocation complexity and management difficulty increase

Engineering Contradiction:
Improveresource utilizationVSAvoidresource allocation complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent applies universality by creating a standardized virtual node architecture with common storage abstraction interfaces that can be deployed across multiple hosts. The persistent storage layer provides universal data access patterns that work consistently regardless of which host the virtual node runs on, simplifying resource allocation management while enabling efficient multi-node deployment on single hosts and across the cluster

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent utilizes parameter changes by dynamically adjusting storage allocation parameters and resource configuration based on virtual node requirements and host capacity. The system monitors and modifies storage I/O parameters, data placement strategies, and resource allocation settings to optimize performance for multiple virtual nodes on a single host, reducing management complexity while maintaining high resource utilization

Inventive Principle:
Principle #35Parameter changes

3Adaptability or versatility

If persistent data is stored remotely rather than locally on each host, then migration capability and system flexibility improve, but data access speed and storage performance may deteriorate

Engineering Contradiction:
Improvemigration capabilityVSAvoiddata access speed
Core Design Contradiction:
Adaptability or versatilityVSSpeed

Solution Approach 1:

The patent applies segmentation by dividing data storage into hot data (frequently accessed) and cold data (less frequently accessed), with different storage strategies for each segment. Hot data is cached in local or high-speed remote storage to maintain fast access speeds, while cold data is stored in cost-effective remote locations. This segmentation enables migration capability through remote storage while preserving data access speed for critical operations

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements nested doll by creating a hierarchical storage architecture where local storage, remote storage, and archive storage are nested in layers. The virtual node first accesses data from the fastest local storage layer, then falls back to remote storage layers as needed. This nested structure provides migration capability through remote storage while maintaining fast data access speeds through the hierarchical lookup and caching mechanism

Inventive Principle:
Principle #7Nested doll (Nesting)

Data Source

PatentEP3647947B1Enhanced data storage of virtual nodes in a data processing environment
Publication Date: 2022.07.13 HEWLETT PACKARD ENTERPRISE DEV LP
  • EP3647947B1 patent drawingFigure 1
  • EP3647947B1 patent drawingFigure 2
  • EP3647947B1 patent drawingFigure 3A

AI summary

Described herein are systems, methods, and software to enhance the management of virtual nodes in a computing environment. In one example, a method includes identifying a transition event for a first version of a virtual node executing on a first host, identifying a second host to support a second version of the virtual node, and initiating execution of the second version of the virtual node on the second host wherein the virtual nodes comprise a non-persistent portion of storage on the respective executing host and a persistent portion of storage on a second storage element.