Network Cache Architecture Using Stripped Packets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current caching methods require caches to understand specific protocols and maintain application-specific state, making them inflexible and costly to upgrade, especially in mobile networks where session state migration and cache distribution are complex, and there is a need for seamless fallback in case of cache failures.
Innovation Solution
Implementing a cache architecture that uses stripped-down packets with pointer information to regenerate full packets, allowing caches to be application-agnostic and eliminating the need for session state migration, with a fallback mechanism to ensure continuous service even when caches are not available.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If caches store entire files from file servers, then file serving speed improves, but cache complexity and protocol understanding requirements increase
Solution Approach 1:
The patent segments the file into multiple data segments distributed across multiple caches. Instead of one cache storing the entire file, the file is divided into segments that can be independently stored and retrieved by different caches, reducing the complexity burden on individual caches while maintaining fast serving speed.
Solution Approach 2:
The patent extracts the indexing and assembly function from individual caches and centralizes it in the file server. The file server maintains the file index and assembles segments from multiple caches, allowing caches to be simpler storage units without requiring full protocol understanding or state maintenance capabilities.
2Reliability
If caches maintain application-specific state, then protocol compatibility improves, but adaptability to new applications decreases
Solution Approach 1:
The patent creates a universal cache architecture where caches store generic data segments without application-specific state. The file server provides multi-functionality by handling different applications (streaming, downloading, etc.) and translating them into uniform segment retrieval operations, allowing the same cache infrastructure to serve multiple applications without modification.
Solution Approach 2:
The file server acts as an intermediary between applications and caches. It translates various application protocols and requirements into standardized segment retrieval requests that simple caches can fulfill, eliminating the need for caches to understand or maintain application-specific state while preserving protocol compatibility.
3Reliability
If session state is migrated between caches during user mobility, then service continuity improves, but system complexity increases
Solution Approach 1:
The patent extracts session state and file indexing from individual caches and centralizes it in the file server. When users move between caches, the file server maintains the session state and coordinates segment retrieval from appropriate caches, eliminating the need for complex state migration between caches while ensuring service continuity.
Solution Approach 2:
The file server serves as a mediator during user mobility events. It receives segment requests from any cache, maintains session continuity, and coordinates with the appropriate caches to fulfill requests, thereby simplifying the mobility management complexity that would otherwise reside in the caches themselves.
4Manufacturing precision
If protocol-specific caches are deployed for each application, then protocol handling accuracy improves, but manufacturing and maintenance costs increase
Solution Approach 1:
The patent deploys universal caches that handle generic data segment storage without application or protocol specificity. The file server provides protocol handling accuracy by translating various protocols into standardized segment operations, eliminating the need to manufacture and maintain multiple protocol-specific cache types while preserving accurate protocol handling.
Solution Approach 2:
The file server acts as an intermediary that ensures protocol handling accuracy. It translates different application protocols into uniform segment retrieval operations that simple universal caches can execute, thereby maintaining protocol accuracy without requiring complex protocol-specific cache implementations.
Data Source
AI summary
There is described a method and apparatus for sending data through one or more packet data networks. A stripped-down packet is sent from a packet sending node towards a cache node, the stripped down packet including in its payload a pointer to a payload data segment stored in a file at the cache node. When the stripped-down packet is received at the cache node, the pointer is used to identify the payload data segment from data stored at the cache node. The payload data segment is inserted into the stripped-down packet in place of the pointer so as to generate a full size packet, which is sent from the cache node towards a client.