Storage Controller Concurrent Cluster Node Operation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage systems are limited by being mutually exclusive, forcing companies to replace entire systems when transitioning from non-clustered to clustered configurations, leading to costly upgrades and potential disruptions to business processes.
Innovation Solution
A storage controller that can operate concurrently as both an individual non-clustered storage server and a node in a storage server cluster, allowing for gradual transitioning of data ownership and system architecture to accommodate changing storage needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a storage system is configured as a clustered system to improve scalability, then the storage capacity and performance can be expanded, but the system complexity and cost of transition increase
Solution Approach 1:
The storage system is segmented into independent storage servers that can operate autonomously as non-clustered units while simultaneously being part of a clustered configuration. Each storage server maintains its own identity and can function independently, yet collectively they form a scalable cluster architecture, allowing gradual transition from non-clustered to clustered operation.
Solution Approach 2:
Storage servers are designed with multi-functionality to operate in dual modes: as individual non-clustered storage servers and as nodes within a clustered storage system. This universal capability allows the same hardware and software infrastructure to support both operational modes, enabling flexible scaling without requiring complete system replacement.
2Adaptability or versatility
If a company transitions from non-clustered to clustered storage systems, then storage needs can be accommodated, but complete system replacement is required leading to costly upgrades
Solution Approach 1:
The system enables preliminary configuration where storage servers are initially deployed as non-clustered units and can later be integrated into a clustered architecture without replacement. This preliminary action allows organizations to start with simpler, lower-cost individual servers and progressively transition to clustered configurations as storage needs evolve, avoiding premature investment in complete clustered system replacement.
Solution Approach 2:
The storage system architecture is designed to be dynamic, allowing storage servers to transition between non-clustered and clustered operational states. This dynamic capability enables organizations to adapt their storage infrastructure to changing needs over time, maintaining existing investments while progressively implementing clustering to accommodate growing storage requirements.
3Reliability
If storage servers are configured as failover partners to increase reliability, then system reliability improves, but the ability to scale and transition to clustered architecture is limited
Solution Approach 1:
The storage system segments reliability functions from the overall architecture, allowing individual storage servers to be configured as failover partners while maintaining the capability to join a clustered architecture. Each server maintains its own failover capabilities independently, yet can simultaneously participate in cluster-wide load balancing and resource sharing, providing both reliability and scalability.
Data Source
AI summary
A storage controller is operable concurrently as an independent storage server and as at least a portion of a node of a storage server cluster. A network storage system comprises multiple storage controllers and multiple independent storage servers, each storage server implemented in a separate storage controller, each of the storage server configured to present to users a separate system image of stored data. The storage servers are independently operable and manageable. The system further includes a storage server cluster that includes cooperating storage server nodes, distributed among the storage controllers. The storage server cluster presents to users a single system image of data stored in the server cluster independently of the node or nodes in which said data resides. The storage server cluster is operable independently of the independent storage servers and is operable concurrently with operation of at least one of the independent storage servers.