Distributed RAID-1 Local Copy Operations for Bandwidth Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed RAID-1 systems face inefficiencies in bulk copy operations due to bandwidth saturation and increased latency, which slow down data transfer and interfere with other operations.
Innovation Solution
Converting distributed copy operations into local copy operations using high-bandwidth, low-latency communication channels at each location, allowing for independent and faster data transfer without consuming bandwidth on long-haul channels, and reverting back to normal mode upon completion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If distributed copy operations are performed using long-haul communication channels, then data can be transferred between geographically separated disks, but the limited bandwidth and increased latency slow down the operation and interfere with other operations
Solution Approach 1:
The patent segments the distributed copy operation into multiple local copy operations. Instead of copying data across geographically separated locations through long-haul channels, the system divides the source and target disks into local groups at different locations, and performs copy operations locally within each location. This segmentation eliminates the need for slow long-haul data transfer while maintaining the distributed RAID-1 architecture.
Solution Approach 2:
The patent introduces local storage controllers as intermediaries at each geographic location. These controllers manage the copy operations locally, acting as mediators that enable data to be copied between source and target disks within the same location without requiring direct communication over long-haul channels. This intermediary approach resolves the bandwidth and latency constraints of distributed copy operations.
2Reliability
If bulk copy operations are performed across distributed RAID-1 volumes, then data redundancy is enhanced, but the communication channels become saturated, slowing down the operation itself
Solution Approach 1:
The patent segments the bulk copy operation into multiple independent local copy operations that can be performed simultaneously at different geographic locations. By dividing the overall copy task into local segments, the system achieves data redundancy enhancement without saturating any single communication channel, as each location uses its own local high-bandwidth channels independently.
Solution Approach 2:
The patent performs copy operations at multiple locations simultaneously, executing more copy operations in parallel than would be possible with sequential long-haul transfers. This partial distribution of the copy task across multiple locations enables the system to achieve complete data redundancy faster by performing excessive local copy actions concurrently rather than waiting for sequential long-haul transfers.
3Adaptability or versatility
If bulk copy operations utilize long-haul communication channels, then distributed data can be copied, but other operations that utilize the channels are interfered with and slowed down
Solution Approach 1:
The patent segments the copy operation to use only local communication channels at each location, isolating the bulk data transfer traffic from long-haul channels. This segmentation allows other operations to continue using long-haul channels without interference, while the distributed copy functionality is achieved through coordinated local operations at multiple segments or locations.
Solution Approach 2:
The patent extracts the data transfer function from the long-haul communication channels and relocates it to local communication channels at each geographic location. By taking out the bulk data transfer requirement from the shared long-haul channels, the system enables distributed copy operations without interfering with other operations that need to use those channels, maintaining channel availability for normal storage operations.
Data Source
AI summary
In a distributed RAID-1 (DR1) copy operation, operation of source and target DR1 volumes is first converted to a local-copy operating mode in which the distributed copy operation is converted into a set of local copy operations. Each t is performed at a respective location of the disks of the DR1 volumes and involves copying data of the source DR1 from a first disk to a second disk via a local second communication channel of relatively high bandwidth. Following conversion, the local copy operations are performed at the locations. Upon completion of the local copy operations, storage operations to the source and target DR1 volumes are temporarily suspended, operation of the source and target DR1 volumes is converted back to the normal operating mode, and storage operations to the source and target DR1 volumes are resumed.