Browse Prior Art Database

Method and Apparatus for Coordinating a Single Active and Multiple Hot-Standby Systems in an Automated FailOver Configuration Disclosure Number: IPCOM000245147D
Publication Date: 2016-Feb-14
Document File: 6 page(s) / 92K

Publishing Venue

The Prior Art Database


Disclosed is a method and apparatus for coordinating a single active and multiple hot-standby systems in an automated service backup and resumption configuration that are dispersed across multiple computer systems incorporated in a single loosely coupled computer complex.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 28% of the total text.

Page 01 of 6

Method and Apparatus for Coordinating a Single Active and Multiple Hot-Standby Systems in an Automated FailOver Configuration

Consider a computing environment in which a program, the Active software unit, provides an ongoing software service. Elsewhere, another equivalent software unit, the Hot-Standby software unit, executes on a different computer system that is loosely coupled

with the computer system that is hosting the Active software unit. In the event of a failure of the Active software unit or its hosting computer system, the Hot-Standby software unit must immediately detect the failure of the Active unit or of its hosting computer system, and assume the Active unit's processing responsibilities in as immediate and seamless a manner as possible. On the other hand, if the Active unit is to be stopped, perhaps for maintenance purposes, it may or may not be desirable for the Hot-Standby unit to assume the Active software unit's processing responsibilities. It also may be desirable to have multiple Hot-Standby software units available, so that the loss of a single Hot-Standby software unit or its host computer system does not compromise the ability to respond quickly to the loss of the Active software unit. In a multiple Hot-Standby configuration, all Hot-Standby software units must react in a coordinated manner to the loss of the Active software unit , electing only one to assume the Active responsibilities.

Typically, this type of problem is resolved by maintaining a communication channel between the Active and Hot -Standby software units. This presents problems of overhead and responsiveness due to polling over the communications channel . It is only as effective as the communication medium being used. Configuration data about the location and name of the different software units must be maintained, and this increases in a non-linear manner as more software units are added.

The proposed solution relies on the availability of a serialization service provided by the computer operating system. The serialization service extends across all computer systems engaged in a loosely coupled configuration. Using logical resources, the serialization system immediately reacts to the loss of the software owner of a logical resource or the loss of the resource owner 's host computer system. The serialization service reclaims the logical resource and assigns ownership to the next requestor waiting for the logical resource. Requests for resource access in this solution are always exclusive. Thus, only one resource requester at a time can have ownership of a logical resource, and any other resource requester for the same logical resource remains in the resource queue (i.e. is enqueued ) on the logical resource.

Serialization Service Interface

The serialization service interface has the following requirements and capabilities.

Resource requesters can query the serialization service for the status of a resource queue. The query response indica...