Distributed State Management in Computing Fabric
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud computing systems face inefficiencies and unreliability due to reliance on external storage for state information, leading to high latencies and potential failures in retrieving data from remote storage.
Innovation Solution
Implementing a computing fabric with distributed storage and replication of state information across multiple nodes, allowing for co-location of software applications and their data, reducing the need for external storage and ensuring high availability and persistence of data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If external storage is used to store state information, then storage capacity is improved, but access latency and reliability deteriorate
Solution Approach 1:
The patent merges storage and computation by co-locating state information with computing nodes within the computing fabric. Instead of separating storage and computation (external storage architecture), the system combines them so that state information resides on the same nodes that execute software applications, eliminating network access latency while maintaining adequate storage capacity through distributed storage across multiple nodes.
Solution Approach 2:
The patent transitions from a centralized external storage architecture to a distributed storage architecture across multiple computing nodes. This dimensional change distributes state information across the computing fabric, transforming the storage model from remote centralized access to local distributed access, thereby reducing latency while preserving storage capacity.
2Quantity of substance
If external storage is used to store state information, then storage capacity is improved, but system reliability deteriorates
Solution Approach 1:
The patent segments the storage function across multiple computing nodes within the computing fabric. Instead of relying on a single external storage system, state information is distributed and replicated across multiple nodes, so that failure of one node does not result in complete data loss or system failure, thereby improving reliability while maintaining storage capacity.
Solution Approach 2:
The patent implements replication of state information across multiple computing nodes. By creating and maintaining copies of state data on different nodes within the computing fabric, the system ensures that if one node fails, other nodes with replicated data can continue operation, thereby improving system reliability without sacrificing storage capacity.
3Reliability
If state information is replicated across multiple nodes, then availability is improved, but storage complexity increases
Solution Approach 1:
The patent makes computing nodes universal by enabling them to perform multiple functions: computation, local storage, and data replication. Each node within the computing fabric serves as both a computing resource and a storage resource, eliminating the need for dedicated external storage infrastructure and simplifying the overall storage architecture while maintaining high availability through replication.
Data Source
AI summary
Various techniques for distributed storage of state information in a computing fabric are disclosed herein. In one embodiment, a method includes executing a software application at a computing device of a computing fabric, and thus resulting in a change to a portion of a copy of state information associated with the software application stored in the computing device. The method also includes automatically replicating the change to the portion of the copy of state information stored on the computing device to one or more other computing devices in the computing fabric. The one or more other computing devices individually contain a copy of the state information associated with the software application.