Memory Replica Replication for Server Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for rapidly cloning a database server with state to a secondary system are resource-intensive and inefficient, particularly in active/active setups, leading to high resource consumption and prolonged spinning up times, and often require database-specific support, limiting scalability and high availability/disaster recovery capabilities.
Innovation Solution
A method that detects changes on a primary computer system's memory device, monitors memory controllers, and triggers a snapshot creation procedure to replicate the complete memory and processor states, allowing for hardware-supported replication and synchronization, enabling rapid and continuous replication of memory changes to an alternate system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a second server is continuously running with replicated storage (active/active setup), then high availability and disaster recovery capabilities are improved, but resource consumption and power consumption increase permanently
Solution Approach 1:
The patent applies preliminary action by continuously replicating storage data to a secondary system in advance, so that when failover is needed, the secondary system is already prepared and can take over immediately. The secondary system remains in a standby state with replicated data but does not need to be fully operational, reducing resource consumption while maintaining high availability capability.
2Adaptability or versatility
If database-specific support is used for replication (e.g., Geographically Dispersed Parallel Sysplex), then active/active setup is enabled, but the solution requires significant database help and cannot be implemented without database-specific protocols
Solution Approach 1:
The patent uses copying by creating a secondary system that replicates the storage data and system state of the primary system. This copy can take over when the primary fails, providing active/active capability. The copying approach is generic and does not require database-specific protocols, as it operates at the storage and system state level rather than requiring database application layer support.
3Ease of operation
If virtualization is used for live guest relocation (e.g., VMware vMotion), then server cloning is enabled, but the method is limited when the server to be relocated is huge
Solution Approach 1:
The patent applies segmentation by dividing the server system into storage data and system state components that can be replicated independently to a secondary system. This allows even large servers to be cloned by breaking down the replication into manageable storage blocks and system state elements, removing the size limitations that constrain virtualization-based approaches.
4Productivity
If storage replication is used to spin up a second server, then cloning is enabled, but the spinning up process takes a long time
Solution Approach 1:
The patent applies preliminary action by continuously replicating storage data and system state to a secondary system in advance. When cloning or failover is needed, the secondary system is already prepared with all necessary data and can be activated immediately, eliminating the long spinning-up time associated with traditional storage replication methods.
Data Source
AI summary
A method for maintaining a memory replica of a primary computer system to be used by an alternate computer system is provided. Provided are detecting changes performed on a memory device of the primary computer system by software executed on the primary computer system. Also provided is monitoring at least one memory controller of the primary computer system; triggering, based on the monitoring, a snapshot creation procedure, the snapshot creation procedure comprising creating a memory snapshot that comprises the detected changes.