Decoupling Primary and Secondary Storage via Intermediate Tier
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage architectures face challenges in scalability, performance, and manageability, particularly in supporting continuous backup and disaster recovery, as they often require significant CPU, memory, and network resources, and introduce latency and disruption due to the direct involvement of primary storage in backup operations.
Innovation Solution
A storage system with a data management tier that decouples primary and secondary storage by using intermediate storage, allowing for off-hosting of backup operations and reducing latency, while maintaining high-performance through local storage and centralized management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If direct attached storage is used for primary storage, then performance is improved, but backup operations cause latency and disruption because the processor is directly involved
Solution Approach 1:
The patent extracts backup operations from the primary storage processor by introducing a separate secondary storage system. The primary storage processor focuses on fast data access while a dedicated backup processor handles backup operations to secondary storage, eliminating the latency and disruption caused by backup operations on the primary storage processor.
Solution Approach 2:
The patent introduces an intermediary storage layer and backup processor that mediates between primary storage and secondary storage. This intermediary layer allows the primary storage processor to maintain high performance while the backup processor independently manages backup operations to secondary storage, preventing direct involvement of the primary processor in backup tasks.
2Speed
If direct attached storage is used, then high performance is achieved, but scalability is limited and resources must be over-provisioned
Solution Approach 1:
The patent creates a universal storage architecture where secondary storage can serve multiple purposes: backup storage, archive storage, and additional primary storage capacity. The system can dynamically allocate resources between different storage functions, improving scalability and resource utilization without sacrificing primary storage performance.
Solution Approach 2:
The patent adds a new dimension to the storage architecture by introducing hierarchical storage layers (primary storage and secondary storage) that operate independently. This dimensional expansion allows the system to scale storage capacity without directly impacting primary storage performance, enabling flexible resource allocation and eliminating the need for over-provisioning.
3Adaptability or versatility
If network attached storage or storage area networks are used, then scalability is improved, but latency increases due to bottleneck phenomena
Solution Approach 1:
The patent applies local quality by providing direct attached primary storage for high-performance data access while using network-attached secondary storage for backup and archive operations. Each storage layer is optimized for its specific function, allowing the system to achieve both low latency for primary operations and high scalability for secondary operations without the bottleneck problems of pure network-attached storage.
Data Source
AI summary
A method of storing data in a storage system is provided. The method includes writing data, from a virtual machine in a compute node, to a primary storage coupled to the virtual machine and sending a first copy of the data from the compute node to a data node. The method includes writing the first copy of the data from the compute node to an intermediate storage coupled as direct attached storage to the data node, and writing a second copy of data from the intermediate storage to a secondary storage, wherein at least one method operation is performed by a processor. A storage system is also provided.