Virtual Storage Array Prefetching WAN Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage systems for enterprises with multiple branch locations face inefficiencies due to the high cost and inefficiency of deploying and maintaining separate data storage at each branch, exacerbated by the slow and latency-prone wide area networks, which prevent effective consolidation of data storage across branches.
Innovation Solution
The implementation of virtual storage arrays that consolidate branch data storage at centralized data centers, using prefetching and caching to overcome bandwidth and latency limitations, and employing quality of service features to optimize performance, allowing branches to access data as if it were locally stored while actually storing it at the data center.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If data storage is consolidated at centralized data centers via WAN, then cost and management efficiency are improved, but access speed and latency worsen
Solution Approach 1:
The system performs preliminary actions by prefetching data blocks before they are actually requested by clients. The storage system anticipates future read requests and retrieves data blocks in advance over the WAN, storing them in local cache memory at branch locations. This preliminary action eliminates the latency penalty when clients actually need the data, as the data is already locally available.
Solution Approach 2:
The invention introduces an intermediary component - a cache memory system at branch locations - that mediates between the centralized data center storage and local clients. This intermediary holds copies of frequently accessed data blocks locally, allowing clients to access data at local speeds while the actual data resides at the centralized data center. The cache acts as a buffer that hides the WAN latency from clients.
2Speed
If separate data storage is deployed at each branch location, then access speed is improved, but cost and management complexity increase
Solution Approach 1:
The system merges the advantages of both distributed and centralized storage by combining local cache memory at branch locations with centralized data center storage. The cache memory is distributed across branches for fast local access, while the actual data storage is centralized at the data center for simplified management. This hybrid approach eliminates the need to manage separate storage systems at each branch while maintaining fast access speeds through local caching.
Solution Approach 2:
The invention uses copying by creating local copies of data blocks in cache memory at branch locations. Instead of storing complete data sets at each branch, the system selectively copies only the data blocks that are likely to be accessed locally. This copying strategy provides fast local access to frequently used data while keeping the primary storage centralized, reducing both cost and management complexity.
3Adaptability or versatility
If excess storage capacity is deployed at each branch, then future growth is accommodated, but cost increases due to unused capacity
Solution Approach 1:
The system introduces dynamics by making storage capacity flexible and on-demand. Instead of deploying fixed excess storage capacity at each branch, the system dynamically allocates storage resources from the centralized data center. The cache memory at each branch is dynamically populated with data blocks based on actual access patterns and needs. This dynamic approach allows branches to access additional storage capacity as needed without having permanently deployed unused capacity.
Solution Approach 2:
The centralized data center storage infrastructure serves multiple branches simultaneously, providing universal storage capacity that can be shared across the entire organization. The same storage resources at the data center can serve any branch that needs them, eliminating the need for each branch to have dedicated excess capacity. The system achieves multi-functionality by allowing the centralized storage to fulfill the storage needs of multiple different branches dynamically.
Data Source
AI summary
Virtual storage arrays consolidate branch data storage at data centers connected via wide area networks. Virtual storage arrays appear to storage clients as local data storage, but actually store data at the data center. Virtual storage arrays may prioritize storage client and prefetching requests for communication over the WAN and/or SAN based on their associated clients, servers, storage clients, and/or applications. A virtual storage array may transfer large data sets from a data center to a branch location while providing branch location users with immediate access to the data set stored at the data center. Virtual storage arrays may be migrated by disabling a virtual storage array interface at a first branch location and then configuring another branch virtual storage array interface at a second branch location to provide its storage clients with access to storage array data stored at the data center.