Continuous Data Protection via I/O Filtering Framework
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing snapshot-based backup and recovery technologies for virtual machines (VMs) result in high recovery point objectives (RPOs) due to the expense and user impact of frequent snapshot taking, typically leading to RPOs in the tens of minutes, which is inadequate for applications requiring near-zero data loss.
Innovation Solution
A system intercepts virtual disk I/Os in real-time, replicating them to a backup site, eliminating the need for frequent snapshots, and forms a recoverable snapshot-log chain by applying the replicated I/O stream on top of a base snapshot, allowing for near-real-time data recovery with an RPO of seconds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If snapshots are taken frequently to reduce RPO, then data recovery precision is improved, but system performance deteriorates and user experience worsens
Solution Approach 1:
The system segments the backup process into two distinct phases: (1) taking a base snapshot periodically with minimal impact, and (2) continuously logging I/O operations asynchronously in the background. This segmentation allows frequent data capture without the performance penalty of frequent full snapshots, resolving the contradiction between RPO improvement and user experience maintenance.
Solution Approach 2:
The system performs preliminary action by taking a base snapshot before continuous logging begins. This preliminary snapshot provides a stable foundation, allowing subsequent continuous I/O logging to proceed without repeatedly freezing the VM for snapshot operations, thus improving RPO while maintaining user experience.
2Measurement precision
If snapshots are taken frequently to reduce RPO, then data recovery precision is improved, but system resource consumption increases
Solution Approach 1:
The backup process is segmented into a resource-intensive base snapshot taken periodically and a resource-light continuous logging phase. This segmentation eliminates the need for frequent full snapshots, significantly reducing system resource consumption while maintaining low RPO through continuous asynchronous I/O logging.
Solution Approach 2:
Instead of creating full snapshot copies frequently, the system copies only the I/O operation logs to the backup site. This copying approach requires minimal system resources compared to full snapshot operations, enabling frequent data capture without proportionally increasing resource consumption.
3Measurement precision
If snapshots are taken frequently to reduce RPO, then data recovery precision is improved, but operation complexity increases
Solution Approach 1:
The system segments snapshot management into a simple periodic base snapshot operation and an automated continuous logging process. This segmentation reduces operation complexity by eliminating the need to manage frequent full snapshots, while the automated logging framework handles continuous data capture without manual intervention.
Solution Approach 2:
The system introduces an intermediary logging mechanism between the VM and the backup site. This logging layer acts as a mediator that continuously captures I/O operations without requiring complex snapshot coordination, simplifying the overall backup management while achieving low RPO.
Data Source
AI summary
Example embodiments relate generally to systems and methods for continuous data protection (CDP) and more specifically to an input and output (I/O) filtering framework and log management system to seek a near-zero recovery point objective (RPO).