Virtual Disk Replication via Parallel Log Files
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current approaches for tracking changes in virtual machines during replication are I/O-intensive, leading to additional storage overhead and impacting response times, especially when using differencing disks, which were not designed for replication and can result in inefficient data transfer and synchronization issues during disasters.
Innovation Solution
The implementation of replication log files that capture changes in parallel with virtual disk updates, allowing for efficient data transfer and reducing storage I/O operations by accumulating write requests in memory before writing to disk, and using a log file format agnostic to virtual hard disk formats to facilitate seamless change tracking and storage migration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If differencing disks are used for tracking changes during replication, then change tracking is enabled, but storage I/O operations increase and response times are impacted
Solution Approach 1:
The patent segments the change tracking function from the virtual disk operations by introducing a separate log file that records changes independently. This allows the virtual disk to be accessed for normal operations while changes are captured in the log file without blocking I/O operations, thus maintaining response time while enabling change tracking.
Solution Approach 2:
The patent introduces a log file as an intermediary mechanism between the virtual disk and the replication process. The log file captures changes asynchronously, serving as a mediator that enables change tracking without directly interfering with the performance-critical virtual disk operations.
2Productivity
If log files are used to capture changes in parallel, then storage I/O operations are reduced, but additional data transfer overhead is introduced
Solution Approach 1:
The patent extracts only the essential change information from the virtual disk operations and stores it in the log file, rather than transferring entire disk blocks. This extraction approach reduces the amount of data that needs to be transferred during replication while maintaining complete change tracking capability.
Solution Approach 2:
The patent changes the parameter being tracked from entire disk blocks to specific change events recorded in the log file. This parameter change allows for more efficient data representation and transfer, as only the necessary change information is captured and transferred rather than large fixed-size blocks.
3Reliability
If virtual disk blocks are transferred for replication, then data consistency is maintained, but network bandwidth is consumed inefficiently
Solution Approach 1:
The patent performs preliminary action by capturing changes in the log file as they occur, rather than waiting to transfer entire virtual disk blocks. This allows for incremental replication of only the changed data, maintaining consistency while reducing network bandwidth consumption compared to block-based transfer methods.
Data Source
AI summary
Techniques involving replication of virtual machines at a target site are described. One representative technique includes an apparatus including a virtual machine configured to provide storage access requests targeting a virtual disk. A storage request processing module is coupled to the virtual machine to receive the storage access requests and update the virtual disk as directed by the storage access requests. A replication management module is coupled to the virtual machine to receive the storage access requests in parallel with the storage request processing module, and to store information associated with the storage access requests in a log file(s). The log file may be transferred to a destination as a recovery replica of at least a portion of the virtual disk.