Browse Prior Art Database

Method and System for saving power in a WPAR environment by using dynamic monitoring and predictive learning Disclosure Number: IPCOM000191650D
Original Publication Date: 2010-Jan-11
Included in the Prior Art Database: 2010-Jan-11
Document File: 2 page(s) / 92K

Publishing Venue



As part of Green Computing, saving power/energy wherever possible has become the new mantra for the industry. IBM is pioneering several technologies to save power. On AIX servers, IBM has introduced WPARs as a server consolidation technology. In this scenario, traditional power saving solutions involve cluster technologies and does not take advantage of the monitoring capabilities of AIX global environment.

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 53% of the total text.

Page 1 of 2

Method and System for saving power in a WPAR environment by using dynamic monitoring and predictive learning

Disclosed is a method and system for saving power in a workload partitioned environment. On AIX Operating System, Workload partitions (WPAR)

provide a secure and isolated environment

for enterprise applications in terms of process, signal and file system space. Software running within the context of Workload Partitions will appear to have its own separate instance of AIX. This capability allows consolidation of workloads. By consolidating, customers can run focused workloads on WPARs. For example, on one AIX Server, customers can run two WPARS. One WPAR can run a Web Server instance and the other WPAR can run a database workload.

Network applications like Web Server are driven primarily by client HTTP (HyperText Transfer

Protocol) requests for a webpage or a dynamic application. When there are no HTTP requests, web server process usually won't do any processing and just listen on the well-defined port.

There is an opportunity to save energy when there are no HTTP client requests to the web server for a longer period of time (for example 3 hours). Hence, in a WPAR environment where one AIX server can run many WPARs, a solution is required to save energy.

The solution works as follows:-
1. 'Monitoring & Control Agent' (runs as a daemon) sits on the global environment and monitors the HTTP requests that are received by the Web Server process running on the WPAR. All the client HTTP request information (including the time on which request is received) are stored in the database.

2. A 'Learning Module' is deployed on the global environment and it contains implementations of predictive modeling algorithm. Learning algorithm is employed to find time periods where there are no active HTTP requests from clients. One such learning algorithm is Naives Bayes Classifier. The learning module then takes a decision to save energy by sending a control code to the 'Monitoring & Control Agent'.

3. The 'Monitoring & Control Agent'...