Replication Splitter ATS Command Interception
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data replication systems face challenges in keeping pace with high data transaction rates at production sites, leading to potential backlogs and system shutdowns, as they struggle to adapt dynamically to changing transaction rates without impacting production site performance.
Innovation Solution
Implementing a method that intercepts atomic test and set (ATS) commands in a replicated environment, where meta data is added to a backlog, sent to a replication appliance, and processed based on success or failure, allowing for dynamic handling of command failures and mode changes to ensure seamless data replication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If journaling is used for continuous data protection, then data can be recovered to any specified point in time, but the backup site cannot keep pace with high data transaction rates at the production site
Solution Approach 1:
The system dynamically adjusts its operation mode based on the backlog status. When the backlog exceeds a threshold, the splitter transitions to Marking on Host (MOH) mode, which marks transactions without immediately replicating them, thereby reducing processing pressure while maintaining protection continuity. This dynamic adaptation allows the system to handle varying data transaction rates without forcing production site shutdown.
Solution Approach 2:
The system performs preliminary actions by marking transactions in advance before they are fully processed. In MOH mode, transactions are marked as protected without immediate replication, allowing the system to prepare for future replication cycles. This preliminary marking enables the backup site to keep pace with production transactions even when replication speed is limited.
2Reliability
If the backup site processes every data transaction synchronously, then data protection is maintained, but the production site must slow down or shut down
Solution Approach 1:
The splitter dynamically switches between normal processing mode and MOH mode based on backlog conditions. When in MOH mode, it decouples the marking of transactions from their replication, allowing the production site to continue at full speed while the backup site processes transactions at its own pace. This dynamic behavior eliminates the need for production site shutdown.
Solution Approach 2:
The splitter acts as an intermediary between the production site and backup site. It intercepts transactions, marks them for protection, and then processes them asynchronously. This intermediary role allows the backup site to protect data without directly constraining production site transaction rates, as the splitter buffers and manages the flow between the two sites.
3Reliability
If the system forces production site shutdown to ensure backup completion, then data protection is ensured, but system downtime increases
Solution Approach 1:
The system performs preliminary marking of transactions before they need to be replicated. By marking transactions in advance and maintaining a backlog of marked transactions, the system ensures that all data is protected without requiring production site shutdown. The MOH mode allows the system to complete backup operations asynchronously during normal production operations.
Solution Approach 2:
The system maintains continuous data protection by operating in MOH mode, where transactions are continuously marked and processed without interruption to production site operations. This continuous operation eliminates the need for periodic shutdowns, ensuring both data protection completion and system availability without time loss.
Data Source
AI summary
In an embodiment, a method and program product is presented for executing a command in a replicated environment comprising a replication appliance and a production site, the replicated environment running in target side processing mode, the method comprising intercepting the command at a splitter reading from the volume, sending the data read from the volume to the replication appliance; and sending the command to volume.