Cloud Infrastructure Pod Segmentation for Scalability Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional cloud infrastructure faces scalability limitations due to architectural constraints of commercially available storage and networking solutions, leading to high costs, performance bottlenecks, and vendor lock-in, which hinder the competitiveness of cloud operators.
Innovation Solution
The implementation of a scalable cloud infrastructure that partitions physical resources into pods within zones, with strong connectivity within pods and weak connectivity between them, using high-speed networks and management server clusters for resource allocation and management, along with DNS translation techniques and image stripes for efficient traffic management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional shared storage and Layer-2 networking are used in cloud infrastructure, then virtual machines can access storage and network resources, but scalability is severely limited and performance bottlenecks occur
Solution Approach 1:
The patent segments the cloud infrastructure into multiple isolated zones, each with its own storage and networking resources. This segmentation allows each zone to operate independently, eliminating the single point of failure and performance bottlenecks associated with shared storage and Layer-2 networking, while enabling linear scalability as new zones can be added without impacting existing ones.
Solution Approach 2:
The patent transitions from traditional shared storage architecture to a distributed storage architecture where storage resources are distributed across multiple zones and accessed via network protocols. This dimensional change from centralized to distributed architecture enables the system to scale horizontally by adding more storage nodes across different zones, removing the scalability limits of shared storage.
2Quantity of substance
If commercially available storage servers are deployed to support large scale cloud deployments, then storage capacity increases, but costs become prohibitively expensive and vendor lock-in occurs
Solution Approach 1:
The patent employs commodity storage servers instead of expensive enterprise-grade storage arrays. These commodity servers can be readily replaced or upgraded without significant capital investment, avoiding vendor lock-in and reducing the cost per unit of storage capacity while maintaining sufficient reliability through redundancy across multiple zones.
Solution Approach 2:
The patent creates a storage architecture where commodity servers perform multiple functions - serving as both compute nodes and storage nodes, and where storage resources can be dynamically allocated to different virtual machines across zones. This universality maximizes the utilization of inexpensive hardware resources.
3Speed
If Layer-2 switching is used to interconnect physical servers, then network performance is optimized, but broadcast traffic increases and scalability is limited
Solution Approach 1:
The patent segments the network into multiple isolated zones that do not participate in each other's broadcast domains. Each zone has its own Layer-2 network, preventing broadcast traffic from propagating across the entire infrastructure. This segmentation maintains optimal network performance within each zone while enabling scalability by allowing new zones to be added without increasing broadcast traffic for existing zones.
4Reliability
If redundant storage elements are deployed to eliminate single points of failure, then reliability improves, but costs increase due to capital outlay
Solution Approach 1:
The patent achieves reliability through architectural segmentation into multiple independent zones, where failure of storage elements in one zone does not affect other zones. This distributed architecture provides inherent fault tolerance without requiring expensive redundant enterprise storage arrays, as each zone can continue operating independently even when others fail.
Data Source
AI summary
A scalable cloud infrastructure serves two or more customers, where each customer is associated with at least one unit of virtual resources. The virtual resources are established by apportioning physical resources in the cloud infrastructure that are partitioned into pods within one or more zones in a scalable manner. Additionally, the cloud infrastructure establishes one or more management server clusters each comprising one or more management servers. The two or more customers create a number of virtual machines within pods in a zone. Due to the scalability of the cloud infrastructure, a console proxy virtual machine and server is introduced to support console access to virtual machines. The console proxy server serves as an intermediary between a browser and a viewed virtual machine configured to maintain viewing session quality while minimizing network impact.