Failover Estimation Using Contradiction Proof
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for estimating failover guarantees in virtualized computing clusters rely on rough approximations, leading to potential false negatives where clusters are capable of recovery but are incorrectly identified as unable to do so.
Innovation Solution
A proof by contradiction method is employed by a failover guaranty estimator module to determine memory usage within working hosts, ensuring that virtual machines in a failure host set can be orphaned without exceeding memory thresholds in other hosts, thereby accurately assessing failover capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the slot-based method is used to estimate failover guarantees, then the estimation process is simple and quick, but the accuracy is low leading to false negative results
Solution Approach 1:
The patent applies proof by contradiction by inverting the estimation approach: instead of directly proving failover capability, it assumes failover is impossible and seeks contradictions. The method assumes a particular host set fails, determines memory usage that would guaranty the largest VM is orphaned, then checks if other VMs force this memory usage - creating a contradiction if they do, thereby proving failover is possible.
Solution Approach 2:
The patent segments the estimation process into distinct steps: (1) assuming a particular host set fails, (2) determining memory usage within working hosts that would guaranty the largest VM is orphaned, (3) checking if other VMs force this memory usage, and (4) concluding based on contradiction. This segmentation makes the complex proof process manageable and systematic.
2Ease of manufacture
If the slot-based method assumes all virtual machines consume the same amount of memory, then the calculation is simplified, but the result becomes a rough approximation that cannot accurately prove cluster recovery capability
Solution Approach 1:
The patent applies local quality by treating different virtual machines differently based on their actual memory characteristics. Instead of using a uniform memory assumption for all VMs, the method identifies the largest VM in the failure host set and calculates memory usage specifically tailored to orphaning that particular VM, then checks if other VMs conflict with this localized memory requirement.
Data Source
AI summary
A failover guaranty estimator module performs a proof by contradiction method showing that a cluster failover guaranty can be met for the cluster. For potential failures for which failover is guaranteed, the method assumes a particular host set of one or more hosts fails, leaving one or more working hosts. The method performs a per-failure host set method for the failure host set. The per-failure host set method determines an amount of memory usage within each working host of the assumed working host set that would guaranty that a largest of the virtual machines in the failure host set would be orphaned. The per-failure host set method determines if the virtual machines in the failure set, other than the largest virtual machine in that set, would force the determined amount of memory usage within working hosts, resulting in, the failover guaranty not being met.