I/O Tracking Splitter for Continuous Data Replication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data protection systems face challenges in maintaining continuous data protection, particularly in keeping pace with high data transaction rates at production sites without causing downtime or data loss, due to the overhead of journaling processes which can lead to a backlog of un-logged transactions at backup sites.
Innovation Solution
A method and system that utilize a splitter with I/O tracking mechanisms to continuously push data from a source side to a target side, including features like initializing and managing I/O tracking mechanisms, writing to delta marking caches, and merging tracking mechanisms to ensure seamless data replication and recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If journaling is used for continuous data protection, then data can be recovered to any point-in-time, but the backup site cannot keep pace with high data transaction rates at the production site
Solution Approach 1:
The patent segments the journaling process into two distinct phases: a fast I/O path for logging transactions and a separate replication path for applying changes. This segmentation allows the production site to maintain high transaction rates while the backup site processes changes at its own pace, resolving the contradiction between recovery capability and processing rate.
Solution Approach 2:
The patent introduces an intermediary mechanism using a protected area and transaction log that mediates between the production site and backup site. The transaction log acts as a buffer, allowing the backup site to catch up with production changes without forcing production to slow down, thus maintaining both high productivity and reliability.
2Reliability
If synchronous backup is used to ensure data integrity, then data protection is improved, but system downtime increases during recovery
Solution Approach 1:
The patent implements preliminary action by continuously logging transactions and maintaining a protected area with undo information in advance. This prepares the backup site with all necessary recovery data before a disaster occurs, enabling rapid recovery without requiring time-consuming synchronous operations during the actual recovery event.
Solution Approach 2:
The patent ensures continuity of useful action by maintaining the journaling and replication processes continuously in the background. The transaction log and protected area are updated continuously without interrupting production operations, so recovery can begin immediately when needed, minimizing downtime while ensuring data integrity.
3Reliability
If the backup site processes every I/O transaction to maintain continuous protection, then data protection is improved, but the backup site creates a backlog of un-logged transactions
Solution Approach 1:
The patent segments the processing paths by separating I/O logging from I/O replication. The fast I/O path handles transaction logging independently, while the replication path processes changes asynchronously. This segmentation prevents backlogs by allowing the backup site to process transactions at its own pace without blocking production I/O operations.
Solution Approach 2:
The patent introduces dynamic adaptability by allowing the backup site to adjust its processing rate based on its capacity. The system dynamically balances the replication load, ensuring continuous protection is maintained without creating backlogs. The protected area and transaction log mechanisms enable the backup site to catch up dynamically with production changes.
Data Source
AI summary
Methods and apparatus for eliminating the need for a complete synchronization due to failure of a data protection appliance in a continuous data protection system having a replication splitter. In one embodiment, a continuous data protection system includes a source side having a source side storage array with a splitter and a data protection appliance, where processing includes initiating a source side splitter session, initializing a first I/O tracking mechanism for the splitter session, and activating the splitter to a source side processing active state to continuously push I/O data from the source side to the target side.


