Browse Prior Art Database

Autonomic Centralized Infrastructure Administration via Distributed, Redundant Synchronization and Monitoring

IP.com Disclosure Number: IPCOM000149926D
Original Publication Date: 2007-Apr-12
Included in the Prior Art Database: 2007-Apr-12
Document File: 3 page(s) / 90K

Publishing Venue

IBM

Abstract

This article describes a methodology for providing centrally administered software systems with autonomic, redundant, and distributed capabilities through a system of monitoring and synchronization. With the ability to provide these features, infrastructure management continues from a single point which maintains data integrity and infrastructure security; however, node failure does not result in an administration outage, instead allowing another node to automatically reconfigure itself to pick up the administration role by means of continual configuration synchronization, administration node health cross-monitoring, cluster paritioning avoidance, and administration access reconfiguration.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 54% of the total text.

Page 1 of 3

Autonomic Centralized Infrastructure Administration via Distributed , Redundant Synchronization and Monitoring

All nodes functioning as administration nodes run a process which monitors the availability of the administration tool (URL, etc), the status of all administration nodes, and also synchronizes the administartion tool configuration data. To maintain the data integrity, the primary node's monitoring process watches for configuration changes and backs them up to a file in a configured location. The currently inactive administration nodes monitor the configured location for updated backup files and, when new files are found, applies the changes to its local configuration. This workflow both maintains data integrity while also facilitating distributed synchronization of the administration configuration data without the need for shared filesystems. Any lag time in the application of the changes in configuration is irrelevant since only one node serves the administration functions at a given time. Nodes or servers that are managed as part of the administration realm do not need to run any special processes, as the autonomic administration infrastructure is masked by a single point of entry, such as a DNS alias or VIP. The same holds true for administrators accessing the administration interface (i.e. URL). For an overall picture of what this autonomic administration architecture looks like, refer to Figure 1.

Figure 1 -- Autonomic, Distributed, and Centralized Administration Infrastructure.

1

[This page contains 1 picture or other non-text object]

Page 2 of 3

In the event that during the monitoring (heartbeat) process, a node cannot reach the other administration...