Distributed VM Image Subsystem for Cloud Instantiation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current infrastructure-as-a-service (IAAS) platforms face challenges in efficiently distributing virtual-machine (VM) images across servers in large distributed computing systems, leading to performance bottlenecks, increased latency, and data-storage constraints, which impact the temporal and computational efficiency of VM instantiation.

Innovation Solution

A distributed VM-image-distribution subsystem is implemented, where VM images are stored and propagated across multiple servers using EAGER and ON_DEMAND image types, with a decentralized storage approach and an orchestrator subsystem to manage distribution, minimizing network overhead and ensuring predictable instantiation times.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If VM images are centrally stored and distributed across the network, then data-storage capacity is optimized, but network overhead increases and latency increases

Engineering Contradiction:

Improvedata-storage capacityVSAvoidlatency

Core Design Contradiction:

Quantity of substanceVSLoss of time

Solution Approach 1:

The patent segments the centralized VM image storage into distributed local caches across multiple servers. Each server maintains a local cache of VM images, dividing the monolithic storage system into smaller distributed units. This reduces network latency by allowing servers to access images locally while maintaining overall system storage capacity through the distributed cache network.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system performs preliminary actions by pre-caching VM images on multiple servers before they are needed for instantiation. The orchestrator proactively distributes image copies to server caches in advance, so that when instantiation is requested, the image is already available locally, eliminating network transfer delays.

Inventive Principle:

Principle #10Preliminary action

2Speed

If VM images are distributed to all servers, then instantiation speed improves, but network overhead and storage redundancy increase

Engineering Contradiction:

Improveinstantiation speedVSAvoidnetwork overhead

Core Design Contradiction:

SpeedVSLoss of energy

Solution Approach 1:

The patent applies local quality by making VM image availability server-specific rather than universally distributed. Each server caches images based on local demand patterns and workload characteristics. This selective local caching optimizes instantiation speed for each server's specific needs while reducing unnecessary network transfers and storage redundancy across the entire system.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system uses partial action by caching only the necessary portions of VM images on each server rather than complete copies. The orchestrator intelligently determines which image segments to distribute to which servers based on instantiation requests, achieving sufficient instantiation speed without the overhead of complete universal distribution.

Inventive Principle:

Principle #16Partial or excessive action

3Device complexity

If a centralized VM-image distribution system is used, then management is simplified, but the system becomes vulnerable to catastrophic failures and bottlenecks

Engineering Contradiction:

Improvemanagement complexityVSAvoidsystem reliability

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The patent segments the centralized distribution architecture into a distributed orchestrator system. Multiple orchestrator instances operate across different servers, eliminating the single point of failure. Each orchestrator manages local cache operations independently, providing fault tolerance while maintaining manageable complexity through modular design.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system implements beforehand cushioning by creating redundant copies of VM images across multiple server caches and orchestrator instances. This redundancy acts as a cushion against catastrophic failures, ensuring that if one orchestrator or storage node fails, other nodes can continue providing service without complete system failure.

Inventive Principle:

Principle #11Beforehand cushioning (Prior cushioning)

4Loss of time

If VM images are cached locally on servers, then instantiation latency is reduced, but data-storage capacity requirements increase

Engineering Contradiction:

Improveinstantiation latencyVSAvoidstorage capacity

Core Design Contradiction:

Loss of timeVSQuantity of substance

Solution Approach 1:

The patent merges the storage functions of multiple servers into a collective distributed cache system. While individual servers require local storage capacity for caching, the aggregate storage capacity of the distributed system equals or exceeds the original centralized storage. This merging approach reduces instantiation latency through local access while maintaining overall storage efficiency through shared resources.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS11057263B2Methods and subsystems that efficiently distribute VM images in distributed computing systems

Publication Date: 2021.07.06 VMWARE INC

AI summary

The current document is directed to methods and systems that efficiently distribute virtual-machine images (“VM images”) among servers within large, distributed-computer-system-implemented IAAS platforms to facilitate temporally and computationally efficient instantiation of virtual machines within the servers. In implementations discussed below, VM images are stored in a distributed fashion throughout one or more distributed computing systems, using several different VM-image-distribution models, in order to balance computational-resource usage, temporal constraints, and other factors and considerations related to VM-image distribution and VM instantiation.