IP Library Granted Patent US 8,296,418
Granted Patent B2
US 8,296,418 · App. 12/356,990 · Granted Oct 23, 2012

Optimized modification of a clustered computer system

Assignee: International Business Machines Corporation
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 8,296,418
App. No.
12/356,990
Granted
Oct 23, 2012
Kind
B2
Abstract

A method, program product, and data processing system are provided for modifying a cluster in a data processing system, the cluster comprising a plurality of nodes. A monitoring unit monitors resource usage over at least one cycle of a periodic timeframe to obtain monitored resource usage data. A forecasting component forecasts resource usage over a future cycle of the periodic timeframe using the monitored resource usage data to obtain forecast resource usage data. An analysis component analyzes the forecast resource usage data using at least a first rule to determine whether a modification of the cluster is necessary and at least a second rule to determine a time window for the modification of the cluster. A modification component modifies the cluster by adding or removing a node to the plurality of nodes in the time window.

Claims (39)

1. A method of modifying a cluster in a data processing system, the cluster comprising a plurality of nodes, the method comprising:

monitoring resource usage over at least one cycle of a periodic timeframe to obtain monitored resource usage data;

forecasting resource usage over a future cycle of the periodic timeframe using the monitored resource usage data to obtain forecast resource usage data;

analysing the forecast resource usage data using at least a first rule to determine whether a modification of the cluster is necessary and at least a second rule to determine a time window for the modification of the cluster; and

performing the modification of the cluster by adding or removing a node to the plurality of nodes in the time window.

2. The method of claim 1 , further comprising the step of providing a suggested timeframe for the modification to a user.

3. The method of claims 1 , wherein the first rule comprises a first threshold value of said resource usage, the modification of the cluster being deemed necessary if a value of the forecast resource usage data exceeds the first threshold.

4. The method of claim 3 , wherein the first rule further comprising a first time period, the modification of the cluster being deemed necessary if the forecast resource usage exceeds the first threshold for at least the first time period.

5. The method of claim 1 , wherein the second rule comprises a second threshold value of the resource usage and a second time period, wherein the time window is a part of the forecast resource usage data where the resource usage is below the second threshold for at least the second time period.

6. The method of claim 1 , wherein the cluster is a shared nothing cluster.

7. A computer program product comprising a tangible computer useable storage device having encoded thereon:

computer executable program code for monitoring resource usage over at least one cycle of a periodic timeframe to obtain monitored resource usage data;

computer executable program code forecasting resource usage over a future cycle of the periodic timeframe using the monitored resource usage data to obtain forecast resource usage data;

computer executable program code analysing the forecast resource usage data using at least a first rule to determine whether a modification of the cluster is necessary and at least a second rule to determine a time window for the modification of the cluster; and

computer executable program code performing the modification of the cluster by adding or removing a node to the plurality of nodes in the time window.

8. A data processing system for modifying a cluster, said cluster comprising a plurality of nodes, the data processing system comprising:

a processor;

a memory operably associated with the processor; and

a program of instruction encoded on the memory and executable by the processor, the program of instruction comprising:

program instructions for monitoring resource usage over at least one cycle of a periodic timeframe to obtain monitored resource usage data;

program instructions for forecasting resource usage over a future cycle of the periodic timeframe using the monitored resource usage data to obtain forecast resource usage data;

program instructions for analyzing the forecast resource usage data using at least a first rule to determine whether a modification of the cluster is necessary and at least a second rule to determine a time window for the modification of the cluster;

program instructions for performing the modification of the cluster by adding or removing a node to the plurality of nodes in the time window.

9. The data processing system of claim 8 , wherein the program of instruction further comprises program instructions for providing a suggested timeframe for the modification to a user.

10. The data processing system of claim 8 , wherein the first rule comprises a first threshold value of the resource usage, the modification of the cluster being deemed necessary if a value of the forecast resource usage data exceeds the first threshold.

11. The data processing system of claim 10 , wherein the first rule further comprising a first time period, the modification of the cluster being deemed necessary if the forecast resource usage exceeds the first threshold for at least the first time period.

12. The data processing system of claim 8 , wherein the second rule comprising a second threshold value of the resource usage and a second time period, wherein the time window is a part of the forecast resource usage data where the resource usage is below the second threshold for at least the second time period.

13. The data processing system of claim 8 , wherein the cluster is a shared nothing cluster.

14. A data processing system for managing a cluster, the cluster comprising a plurality of nodes, the data processing system comprising:

a monitoring component for monitoring a resource usage over at least one cycle of a periodic timeframe to obtain monitored resource usage data;

a forecasting component for forecasting resource usage over a future cycle of the periodic timeframe using the monitored resource usage data to obtain forecast resource usage data;

storage having stored thereon at least a first rule and a second rule;

an analysis component for analyzing the forecast resource usage data using the first rule to determine whether a modification of the cluster is necessary and the second rule to determine a time window for the modification of cluster;

a modification component for performing the modification of the cluster by adding or removing a node to the plurality of nodes in the time window.

15. The data processing system of claim 14 , further comprising a user interface for providing a suggested timeframe for the modification to a user.

16. The data processing system of claim 14 , wherein the first rule comprises a first threshold value of the resource usage, the modification of the cluster being deemed necessary if a value of the forecast resource usage data exceeds the first threshold.

17. The data processing system of claim 16 , wherein said first rule further comprising a first time period, the modification of the cluster being deemed necessary if the forecast resource usage exceeds the first threshold for at least the first time period.

18. The data processing system of claim 14 , wherein the second rule comprising a second threshold value of the resource usage and a second time period, wherein the time window is a part of the forecast resource usage data where the resource usage is below the second threshold for at least the second time period.

19. The data processing system claim 14 , wherein the cluster being a shared nothing cluster.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 17, 2009
From: ALLWEIL, CLAUS; HEIMES, FRANK
To: INTERNATIONAL BUSINESS MACHINES CORPORATION
Reel/Frame 022268/0816 →
Priority Claims (1)
DE 08150441 · Jan 21, 2008 · national
Continuity (1)
Related Publication 20090187606A1 · Jul 23, 2009