Soft Reservation for Virtualized Workload Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In virtualized computing environments, data loss and excess compute time occur due to the lack of available systems that meet the resource requirements for workload failover during high usage periods, leading to performance detriments.
Innovation Solution
Implementing a soft reservation system that tracks resource requirements and health characteristics of systems to dynamically create reservations on alternative systems, allowing for efficient failover without unnecessary resource reservation, thereby maximizing resource utilization and reducing performance detriments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual network environment configuration is used, then user control is maintained, but data loss and excess compute time occur during failover
Solution Approach 1:
The system performs preliminary actions by proactively monitoring system health characteristics and creating soft reservations on alternative systems before failures occur. When a system exhibits signs of potential failure, the workload is pre-positioned on alternative systems, eliminating the need for time-consuming failover operations when actual failures occur.
Solution Approach 2:
The system implements continuous feedback loops by monitoring health characteristics of systems and dynamically adjusting resource allocation. This feedback mechanism enables the system to detect potential failures early and automatically trigger failover preparations, reducing both data loss and compute time during actual failures.
2Reliability
If resources are reserved for failover, then system reliability improves, but resource utilization decreases
Solution Approach 1:
The system applies dynamics by making resource reservations flexible and adaptive rather than static. Soft reservations are created dynamically based on monitored system health characteristics and can be adjusted or released in real-time, allowing the system to maintain failover capability while maximizing resource utilization during normal operation.
Solution Approach 2:
The system changes the parameter of resource reservation from hard, fixed allocations to soft, dynamic reservations. This parameter change allows resources to be reserved only when needed based on system conditions, improving both reliability and productivity by eliminating unnecessary resource consumption during normal operation.
3Reliability
If hard reservations are created, then failover is guaranteed, but resource flexibility is reduced
Solution Approach 1:
The system applies partial action by creating soft reservations that provide sufficient failover capability without over-reserving resources. This approach ensures adequate protection against failures while leaving remaining resources available for other productive uses, maintaining both reliability and adaptability.
Solution Approach 2:
The system changes the reservation parameter from hard (rigid) to soft (flexible), allowing reservations to be adjusted based on actual system needs and conditions. This parameter change maintains failover guarantees while preserving resource flexibility for adaptive allocation to different workloads.
Data Source
AI summary
A computer program product includes a computer readable medium having computer readable program instructions configured to cause a processor to: track resource requirements for a workload deployed to a virtual environment; monitor one or more health characteristics of one or more systems of the virtual environment; determine whether one or more soft reservations corresponding to the workload should be created on one or more of the systems of the virtual environment; and in response to determining the one or more soft reservations corresponding to the workload should be created, creating the one or more soft reservations on the one or more of the systems of the virtual environment. The one or more of the systems of the virtual environment on which the one or more soft reservations are created are preferably different systems than the system to which the workload was originally deployed. Corresponding systems and methods are also disclosed.