Diskless Client Boot via Multi-SAN Cascade Reads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Storage Area Networks (SANs) are limited in handling random data access requests and simultaneous access by multiple clients, leading to inefficient data transfer and potential bottlenecks due to fixed connections and lack of methods for striped data access across multiple SANs.
Innovation Solution
Establishing multiple iSCSI sessions across multiple SANs to perform cascade reads, allowing a client to determine and allocate read requests to the most efficient SANs for quicker data retrieval and improved boot times by using a connection module and boot module that assesses response efficiency and dynamically reallocates requests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a client establishes a fixed connection to a single SAN, then routing performance is improved and connection stability is maintained, but the client cannot access data from multiple SANs simultaneously and is limited by the performance of a single SAN
Solution Approach 1:
The patent segments the data access process by dividing the operating system into multiple data portions and distributing them across multiple SANs. Each SAN stores a specific portion of the operating system data, allowing the client to establish separate connections to multiple SANs simultaneously and retrieve different portions in parallel, thereby improving adaptability without overwhelming complexity
Solution Approach 2:
The patent implements dynamic connection management where the client can dynamically establish, monitor, and terminate connections to multiple SANs based on real-time performance conditions. The system dynamically allocates data portions to different SANs and can switch between SANs if performance degrades, providing flexibility while managing complexity through automated decision-making
2Productivity
If SANs use fixed MAC address binding to improve routing performance, then connection stability is enhanced, but the system cannot efficiently handle random data access requests or simultaneous access by multiple clients
Solution Approach 1:
The patent segments data access operations into separate read requests for different data portions, each directed to different SANs. This allows the system to maintain stable fixed connections for each SAN while simultaneously improving productivity by parallelizing data retrieval operations across multiple SANs, thereby resolving the contradiction between connection stability and data transfer efficiency
Solution Approach 2:
The patent transitions from a single-dimensional connection model to a multi-dimensional access model where multiple clients can simultaneously access different SANs through different network paths. This dimensional expansion allows the system to handle random data access requests and simultaneous client access without compromising the stability of individual SAN connections
3Ease of operation
If a client is assigned to a specific SAN, then connection management is simplified, but the client may be assigned to a slow SAN even when faster SANs are available, reducing overall system efficiency
Solution Approach 1:
The patent implements feedback mechanisms where the client monitors the performance of each SAN connection in real-time. Based on this feedback, the client dynamically adjusts data allocation decisions, redirecting read requests to faster SANs when available and maintaining connections to slower SANs only when necessary, thereby improving data retrieval speed while keeping operation simple through automated performance-based routing
Solution Approach 2:
The patent changes the assignment parameter from static to dynamic, allowing the system to adjust SAN selection based on real-time performance parameters. The client evaluates SAN performance parameters such as response time and throughput, and uses these changing parameters to make real-time decisions about which SANs to access for specific data portions, maximizing productivity without complicating the operation
Data Source
AI summary
An apparatus, system, and method are disclosed for remotely booting a client from a storage area network (“SAN”). A connection module enables a client, such as a diskless client, to connect to two or more storage area networks (“SANs”), the SANs belonging to a group of redundant SANs, each SAN in the group redundantly storing at least a portion of substantially identical operating system data for the client. The boot module enables the client to remotely boot an operating system from the two or more redundant SANs. The boot module makes at least one read request to each of the two or more connected SANs, each read request configured to retrieve a disparate portion of the operating system data for loading the operating system onto the client. The boot module loads the operating system onto the client using a combination of data retrieved from the two or more connected SANs.