Browse Prior Art Database

Method of sharing Node status in a cluster through Hypervisor.

IP.com Disclosure Number: IPCOM000249191D
Publication Date: 2017-Feb-09
Document File: 4 page(s) / 49K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is the method of sharing the Failure Node Status in a Clustered environment via Hypervisor in which the Failed Node will inform the hypervisor with its cluster information which in turn (hypervisor) broadcasts to all other nodes within the cluster about the node outage information. This prevents the cluster being outage and thus helps in other nodes in the cluster to take over the cluster control and resources to maintain the cluster active.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

1

Method of sharing Node status in a cluster through Hypervisor .

Problem:- In cluster environment, nodes interact with other nodes through Tie breaker disk lock to maintain application resources in cluster. At any given point of time only one node will be holding Tie breaker disc lock and the same node will be maintaining the application resources in active state. It is called as primary node for reference here. The application resources will be moved to other node when the primary node fails by acquiring Tie breaker disc lock. There are situations like primary node crashes by holding Tie breaker lock without releasing it to other nodes to move the application resources. It causes the complete cluster outage and all the application resource will go in inactive state.

The nodes in the cluster should get notify when primary node crashes. The primary node status information helps other nodes to preempt the Tie breaker disk lock and acquire all the application resources. This can be implemented by maintaining cluster information at Hypervisor layer and Hypervisor broadcasting primary node failure information to all other nodes in cluster. Hypervisor will be having a pre-defined network to communicate with other nodes in the cluster.

Features of the solution:

1. Method of maintaining Cluster information statically at Hypervisor layer. 2. Method of broadcasting the failure node information to other nodes in cluster through Hypervisor.

Implementation:

The notification mechanism can be implemented with the help of Hypervisor.

1) Hypervisor will be assigned with predefined IP address to communicate with other nodes in cluster.

2) During cluster creation, primary node notifies Hypervisor about cluster information through H-call.

3) Hypervisor maintains the cluster information in form of a table with node name and IP address statically.

4) Whenever t...