Cluster availability model
First Claim
1. A computer program product comprising a computer useable medium having computer readable code embodied therein for modeling the availability of a cluster, the cluster having a plurality of software components and at least one node, the computer program product adapted when run on a computer to effect steps including:
- determining a plurality of component availability models using a repair model and a plurality of failure parameters, each of the plurality of component availability models corresponding to one of the plurality of software components, wherein the plurality of failure parameters include a failure rate, repair rate and efficacy;
combining the plurality of component availability models, wherein the combining step further comprises obtaining aggregate failure rates, aggregate repair rates, and aggregate efficacies for the plurality of component availability models and wherein the aggregate failure rates, the aggregate repair rates and the aggregate efficacies are obtained for each repair mode in the repair model;
determining repair rates for node and cluster reboots; and
constructing an availability model based on the repair rates and the combined plurality of component availability models;
wherein for each repair mode in the repair model, an aggregate failure rate is a sum of failure rates of the plurality of software components for the repair mode,wherein for each repair mode in the repair model, an aggregate repair rate is a weighted average of repair rates of the plurality of software components for the repair mode, weights being corresponding failure rates of the plurality of software components for the repair mode, andwherein for each repair mode in the repair model, an aggregate efficacy is an weighted average of efficacies of the plurality of software components for the repair mode, weights being corresponding failure rates of the plurality of software components for the repair mode.
2 Assignments
0 Petitions
Accused Products
Abstract
The present invention describes a method and system for creating a cluster availability model that takes into account availabilities of software components in the cluster. Aspects of the disclosure include defining a repair model and failure parameters for a repair mode in the repair model, and modeling availabilities of software components based on the repair mode and failure parameters. Additional aspects include combining availability models of software components in a computationally feasible manner and determining repair rates of node and cluster reboots. A cluster availability model is determined using the combined availability models of software components and the repair rates of node and cluster reboots.
40 Citations
28 Claims
-
1. A computer program product comprising a computer useable medium having computer readable code embodied therein for modeling the availability of a cluster, the cluster having a plurality of software components and at least one node, the computer program product adapted when run on a computer to effect steps including:
-
determining a plurality of component availability models using a repair model and a plurality of failure parameters, each of the plurality of component availability models corresponding to one of the plurality of software components, wherein the plurality of failure parameters include a failure rate, repair rate and efficacy; combining the plurality of component availability models, wherein the combining step further comprises obtaining aggregate failure rates, aggregate repair rates, and aggregate efficacies for the plurality of component availability models and wherein the aggregate failure rates, the aggregate repair rates and the aggregate efficacies are obtained for each repair mode in the repair model; determining repair rates for node and cluster reboots; and constructing an availability model based on the repair rates and the combined plurality of component availability models; wherein for each repair mode in the repair model, an aggregate failure rate is a sum of failure rates of the plurality of software components for the repair mode, wherein for each repair mode in the repair model, an aggregate repair rate is a weighted average of repair rates of the plurality of software components for the repair mode, weights being corresponding failure rates of the plurality of software components for the repair mode, and wherein for each repair mode in the repair model, an aggregate efficacy is an weighted average of efficacies of the plurality of software components for the repair mode, weights being corresponding failure rates of the plurality of software components for the repair mode. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12)
-
-
13. A system for modeling the availability of a cluster, the cluster having a plurality of software components and at least one node, the system comprising:
-
means for determining a plurality of component availability models using a repair model and a plurality of failure parameters, each of the plurality of component availability models corresponding to one the plurality of software components, wherein the plurality of failure parameters include a failure rate, repair rate and efficacy; means for combining the plurality of component availability models, wherein the combining means comprises means for obtaining aggregate failure rates, aggregate repair rates, and aggregate efficacies for the plurality of component availability models and wherein the aggregate failure rates, the aggregate repair rates and the aggregate efficacies are obtained for each repair mode in the repair model; means for determining repair rates for node and cluster reboots; and means for constructing an availability model based on the repair rates and the combined plurality of component availability models; wherein for each repair mode in the repair model, an aggregate failure rate is a sum of failure rates of the plurality of software components for the repair mode, wherein for each repair mode in the repair model, an aggregate repair rate is a weighted average of repair rates of the plurality of software components for the repair mode, weights being corresponding failure rates of the plurality of software components for the repair mode, and wherein for each repair mode in the repair model, an aggregate efficacy is a weighted average of efficacies of the plurality of software components for the repair mode, weights being corresponding failure rates of the plurality of software components for the repair mode. - View Dependent Claims (14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24)
-
-
25. A computer program product comprising a computer useable medium having computer readable code embodied therein for modeling the availability of a cluster, the cluster having a plurality of software components and at least one node, the computer program product adapted when run on a computer to effect steps including:
-
specifying a repair model, the repair model having one or more repair modes; specifying a plurality of failure parameters, for each software component in the plurality of software components, assigning values to the plurality of failure parameters for each appropriate repair mode for the software component, constructing an escalation graph for each of the plurality of software components; combining values of the plurality of failure parameters of the plurality of software components for each repair mode in the repair model; determining repair rates for node and cluster reboots; and constructing an availability model based on the repair rates and the combined plurality of failure parameters. - View Dependent Claims (26, 27, 28)
-
Specification