Failover Processing for Heterogeneous Cluster Nodes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current scale-out approaches for improving system performance and capacity, such as using physical or virtual device clusters, face challenges like long fault recovery durations and lack of core cluster architecture features like failover and load balancing, especially when dealing with heterogeneous nodes.
Innovation Solution
A method and apparatus for failover that migrates data and services between heterogeneous nodes in a cluster, allowing seamless application migration and reducing fault recovery and service interruption times by determining application failures and transferring data and services from one node to another within the cluster.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data and services are migrated between heterogeneous nodes in a cluster, then system scalability and flexibility are improved, but system complexity increases
Solution Approach 1:
The patent implements a universal failover mechanism that works across heterogeneous nodes (different hardware platforms, operating systems, and virtualization environments). The cluster management system provides multi-functional capabilities including failure detection, data migration, service relocation, and load balancing that work uniformly across diverse node types, enabling scalability without proportionally increasing system complexity
Solution Approach 2:
The patent introduces a cluster management system as an intermediary layer between heterogeneous nodes. This mediator handles the complexity of cross-platform communication, data migration, and failover coordination, allowing individual nodes to remain relatively simple while the system as a whole achieves high scalability and flexibility
2Reliability
If failover capability is implemented in a cluster, then service continuity is improved, but system complexity increases
Solution Approach 1:
The patent implements preliminary failure detection mechanisms that continuously monitor node health status before actual failures occur. The cluster management system pre-configures failover policies, data replication strategies, and service migration paths in advance, so that when a failure occurs, the system can execute pre-planned recovery actions without complex real-time decision-making, thus improving service continuity while managing complexity
Solution Approach 2:
The patent employs data copying and replication mechanisms where critical data and services are maintained as copies across multiple nodes. When a node fails, the system can quickly switch to using copied data from a different node, ensuring service continuity. This copying approach simplifies failover logic compared to complex data reconstruction processes
3Adaptability or versatility
If heterogeneous nodes are used in a cluster, then system versatility is improved, but data migration difficulty increases
Solution Approach 1:
The patent implements a data migration system that automatically adapts to different node types by changing migration parameters such as data format, transfer protocol, and storage layout. The cluster management system detects the characteristics of source and destination nodes and adjusts migration parameters accordingly, enabling seamless data migration between heterogeneous platforms without manual configuration or complex conversion processes
Data Source
AI summary
Embodiments of the present disclosure provide a method and apparatus for failover. In an embodiment is provided a method implemented at a first node in a cluster comprising a plurality of heterogeneous nodes. The method comprises: determining whether an application at a second node in the cluster is failed; and in response to determining that the application is failed, causing migration of data and services associated with the application from the second node to a third node in the cluster, the migration involving at least one node heterogeneous to the second node in the cluster. The present disclosure further provides a method implemented at the third node in the cluster and corresponding devices and computer program products.