Fencing management in clusters
View Patent ↗Apparatus, systems, and methods may operate to detect a failure in a failed one of a plurality of nodes included in a cluster, and to fence a portion of the plurality of nodes, including the failed one. Membership in the portion may be determined according to an aggregated value of weighted values assigned to resources and/or services associated with the cluster. Additional apparatus, systems, and methods are disclosed.
1. An apparatus, comprising:
a network node including a failure detection module to detect a failure in a failed node of a plurality of nodes included in a cluster; and
a fencing module to fence a portion of the plurality of nodes, including the failed node and at least one other node, wherein membership in the portion of the cluster is determined according to an aggregated value of weighted values respectively assigned to at least one of resources or services associated with the failed node and the at least one other node within the portion of the plurality of nodes included in the cluster.
2. The apparatus of claim 1 , wherein the network node is included in the cluster.
3. The apparatus of claim 1 , further comprising:
a weight assignment module included in the cluster to assign the weighted values.
4. An apparatus, comprising:
a first network node in a cluster including a failure detection module to detect a failure in a failed node of a plurality of nodes included in a cluster; and
a second network node in the cluster including a fencing module to fence a portion of the plurality of nodes, including the failed node and at least one other node, wherein membership in the portion of the cluster is determined according to an aggregated value of weighted values respectively assigned to at least one of resources or services associated with the failed node and the at least one other node within the portion of the plurality of nodes included in the cluster.
5. The apparatus of claim 4 , further comprising:
at least one of a processing unit or a storage unit as some of the resources.
6. The apparatus of claim 4 , further comprising:
at least one of a network interface or a storage interface as some of the resources.
7. A processor-implemented method to execute on one or more processors that perform the method, comprising:
detecting a failure in a failed node of a plurality of nodes included in a cluster; and
fencing a portion of the plurality of nodes, including the failed node and at least one other node, wherein membership in the portion of the cluster is determined according to an aggregated value of weighted values respectively assigned to at least one of resources or services associated with the failed node and the at least one other node within the portion of the plurality of nodes included in the cluster.
8. The method of claim 7 , further comprising:
assigning a first portion of the weighted values based on relative resource value, and a second portion of the weighted values based on relative service value.
9. The method of claim 7 , further comprising:
assigning a portion of the weighted values that depends on a relative number of the resources and/or the services that exist at a time of assignment, or forecast to exist at a future time.
10. The method of claim 7 , further comprising:
assigning a portion of the weighted values that depends on a ratio of a number of the resources to a number of the services.
11. The method of claim 7 , further comprising:
assigning a portion of the weighted values dynamically, based on at least one of client identity or a time.
12. The method of claim 7 , further comprising:
assigning a portion of the weighted values according to an administrator policy or a pre-selected resource and/or service configuration.
13. The method of claim 7 , further comprising:
assigning some of the weighted values as default values to resources and/or services that do not otherwise have an assigned value.
14. The method of claim 7 , wherein the resources comprise at least one of a processing unit, storage, a storage interface, a network interface, or a floating internet protocol address within the cluster.
15. The method of claim 7 , wherein the services comprise at least one of a database, a web site hosting service, a file access service, or a domain name resolution service.
16. The method of claim 7 , further comprising:
determining the aggregated value as a deterministic function of the weighted values.
17. The method of claim 7 , further comprising:
determining the aggregated value as a function of a first portion of the weighted values associated with the resources, and a second portion of the weighted values associated with the services.
18. A processor-implemented method to execute on one or more processors that perform the method, comprising:
respectively assigning weighted values to at least one of resources or services associated with a plurality of nodes in a cluster; and
when a failure in a failed node of the plurality of nodes is detected, fencing a first portion of the cluster that includes the failed node and at least one other node, the first portion having a lower aggregated value of the weighted values than a second portion of the cluster that is not fenced.
19. The method of claim 18 , wherein the assigning further comprises:
assigning the weighted values to the resources according to a resource type.
20. The method of claim 18 , wherein the assigning further comprises:
assigning the weighted values to the services according to a service type.
21. The method of claim 18 , further comprising:
re-assigning some of the weighted values in response to at least one of an event occurring within the cluster, an event occurring outside the cluster, or a time.
22. The method of claim 18 , further comprising:
determining the aggregated value as a function of the weighted values assigned only to the resources or the services.
23. The method of claim 18 , further comprising:
detecting existence of the failed one as a result of communicating with a cluster membership module.