Storage Session Management in IP-SAN Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In IP-SAN, failure notifications are excessively broadcast, leading to increased network and user loads, as conventional methods lack efficient management of static and dynamic iSCSI session information, resulting in unnecessary burden on servers and networks.
Innovation Solution
A storage session management system with a centralized server that integrates and manages constitutional and operational information of network devices, allowing for targeted failure notifications by grouping and consolidating iSCSI session information, thereby restricting the notification range to affected areas only.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If failure notification is broadcast to all statically connected servers in IP-SAN, then all servers are notified of the failure, but the network load and user burden are excessively increased
Solution Approach 1:
The patent segments the notification scope by dividing servers into different groups based on their access relationships with storage devices. Instead of notifying all servers, only servers in the affected group are notified. This segmentation resolves the contradiction by maintaining reliable notification coverage for affected servers while reducing overall network load by excluding unaffected servers from the notification process.
Solution Approach 2:
The patent applies local quality by making the notification scope dynamic and location-specific rather than uniform across the entire network. The notification is localized to only those servers that have active iSCSI sessions with the failed storage device, determined by checking session information in the session management table. This resolves the contradiction by ensuring reliable notification where needed while minimizing network load elsewhere.
2Reliability
If failure notification is broadcast to all statically connected servers, then no notification is lost, but servers not accessing the failed device are unnecessarily burdened
Solution Approach 1:
The patent introduces dynamics by transitioning from static connection-based notification to dynamic session-based notification. The system continuously monitors iSCSI session states and adjusts the notification scope in real-time based on actual access relationships. This resolves the contradiction by ensuring reliable notification delivery to servers that actually need it while eliminating unnecessary burden on servers with no active sessions to the failed device.
Solution Approach 2:
The patent implements feedback mechanisms by periodically collecting and updating session information from storage devices and comparing it with the session management table. This feedback loop enables the system to accurately determine which servers are actually affected by failures, resolving the contradiction by ensuring notification reliability for affected servers while preventing unnecessary burden on unaffected servers through accurate, real-time session state monitoring.
3Adaptability or versatility
If IP-SAN network is expanded to include more servers, then network extensibility is improved, but the range of failure notification becomes excessively wide
Solution Approach 1:
The patent applies segmentation by organizing servers into distinct groups based on their storage device access patterns. As the IP-SAN network expands, new servers are automatically assigned to appropriate groups based on their session relationships with storage devices. This resolves the contradiction by maintaining network extensibility while preventing notification scope from expanding uniformly, as only servers in the affected group are notified regardless of total network size.
Solution Approach 2:
The patent implements local quality by making notification scope adaptive to local access relationships rather than global network topology. Each storage device and its accessing servers form a local notification domain. This resolves the contradiction by allowing the IP-SAN network to expand freely while keeping failure notification scope localized to only those servers with active sessions, preventing notification scope from becoming excessively wide as the network grows.
Data Source
AI summary
In a storage session management system in a storage area network, the session information blocks periodically collected from the storage devices in the network are grouped based on the information indicating the relation between sessions. That is, since a storage session management server in the management system integrally manages constitution information and operation information, when there is a possibility that circumstances such as a failure and the deterioration of performance occur in information devices, a failure notification module displays such circumstances and makes notification of them to a minimum necessary range that is consolidated by a session consolidation module, thereby a storage session can be managed without applying loads on the respective information devices and the network.