Cluster Workload Pre-Assignment for Split-Cluster Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current failover processes in cluster environments do not address workload duplication across multiple sub-clusters when network connectivity fails between cluster nodes, leading to data inconsistencies and corruption.
Innovation Solution
A computer-implemented method that identifies state changes among cluster resources, predicts resource requirements for queued workloads, determines pre-assignments to sub-clusters based on these requirements, and marks the workloads for scheduling on the assigned sub-clusters, thereby avoiding duplicate deployments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If current failover processes are used to switch primary nodes when network connectivity fails, then node failure tolerance is improved, but workload duplication occurs across multiple sub-clusters
Solution Approach 1:
The patent applies preliminary action by pre-assigning pending workloads to specific sub-clusters before network connectivity fails. When a state change is detected, the system predicts resource requirements and determines pre-assignments of queued workloads to sub-clusters, marking them accordingly. This prevents workload duplication during failover because each sub-cluster knows which workloads it should execute, eliminating the need for random or duplicate deployments across multiple sub-clusters.
2Reliability
If workloads are deployed to multiple sub-clusters during network failure, then availability is maintained, but data consistency deteriorates
Solution Approach 1:
The system performs preliminary assignment of workloads to specific sub-clusters before network partition occurs. By marking workloads with their assigned sub-cluster identifiers, the system ensures that even when network connectivity fails and multiple sub-clusters operate independently, each sub-cluster executes only its assigned workloads. This maintains data consistency while preserving availability, as no duplicate workload execution occurs across different sub-clusters.
3Ease of manufacture
If failover processes only address primary node switching, then node replacement is simplified, but workload management becomes problematic
Solution Approach 1:
The patent extends the simple node-switching approach by adding preliminary workload pre-assignment that automatically occurs when state changes are detected. The system monitors cluster resources, predicts resource requirements for pending workloads, and determines pre-assignments to sub-clusters. This maintains the simplicity of node switching while automatically managing workload distribution, eliminating the need for complex manual workload management during failover events.
Data Source
AI summary
A computer implemented method avoids workload duplication in a cluster environment. The computer identifies a state change among a set of cluster resources in a cluster of nodes. Responsive to identifying the state change, the computer predicts resource requirements for a queued workload. The computer determines a pre-assignment of the queued workload to a sub-cluster according to the resource requirements that were predicted for the queued workload. The computer marks the queued workload to indicate the pre-assignment to the sub-cluster.