Cluster Manager Backup Coordination in Failover Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In failover clustering systems with shared storage, performing backups and restorations is increasingly difficult due to the complexity of determining the appropriate data node for data protection tasks, leading to inefficiencies and potential security issues in traditional federated backup approaches.
Innovation Solution
A method where a cluster manager determines a preferred data node upfront for backup tasks through periodic discovery or pre-backup steps, ensuring that only the most suited data node performs the backup and minimizes inter-node communication, thereby simplifying the backup process and enhancing security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional federated backup approaches are used in failover clustering systems with shared storage, then backup coverage can be achieved across multiple nodes, but the complexity of determining the appropriate data node increases and inter-node signaling becomes excessive
Solution Approach 1:
The patent introduces a cluster manager as an intermediary component that centralizes the backup coordination function. The cluster manager receives backup requests, determines the active data node, and directs backup operations, eliminating the need for complex peer-to-peer node determination and reducing inter-node signaling complexity while maintaining comprehensive backup coverage
Solution Approach 2:
The patent segments the backup function into distinct roles: the cluster manager handles coordination and node determination, while the data node executes the actual backup. This functional segmentation simplifies the overall system by separating control logic from execution logic, reducing the complexity each component must handle
2Reliability
If traditional federated backup approaches are used in failover clustering systems, then backup operations can be performed across nodes, but inter-node signaling becomes excessive and inefficient
Solution Approach 1:
The cluster manager serves as a central intermediary that consolidates communication paths. Instead of multiple nodes exchanging signals with each other to determine backup responsibilities, all backup coordination occurs through the cluster manager, significantly reducing the number of signaling interactions and time required
Solution Approach 2:
The system performs preliminary node determination by maintaining knowledge of which data node is currently active for each asset. This pre-established information allows the cluster manager to immediately direct backup operations without requiring real-time signaling between nodes to determine who should perform the backup
Data Source
AI summary
One or more embodiments of the invention have a cluster manager, or other related component, determine upfront which data node should perform a backup of the asset. This is achieved by running a pre-backup step, either periodically or prior to the reception of a backup request, to determine the preferred node for backup. Once the backup request is received, the cluster manager queries the preferred node, if the preferred node is still the best data node to perform the backup, then that data node performs the backup. If, however, the preferred data node no-longer is the best data node for performing the backup, the current preferred data node will signal the cluster manager with an alternative data node that is best data node to perform the backup and the cluster manager will have that alternative data node perform the backup.