ENERGY-EFFICIENT SERVER LOCATION DETERMINATION FOR CONFIGURATION CHANGES

• US 20110078290A1
• Filed: 09/25/2009
• Published: 03/31/2011
• Est. Priority Date: 09/25/2009
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Verma, Dinesh C., Calo, Seraphin B., Kaminsky, David L., Wang, Xiping
• Granted Patent: US 8,140,652 B2
• Time in Patent Office: Days
• US Class Current: 709/221
• CPC Class Codes: G06F 9/5094 where the allocation takes into account power or heat criteria power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

At a data center controller, a first server location is determined from a set of available server locations at which to change a server active state for a data center. A time and configuration change for the data center are forecasted. A second server location is determined from the set of available server locations at which to change the server active state based upon the forecasted configuration change for the data center. An available server location is selected at which to change the server active state based upon the first server location, the second server location, and the forecasted time associated with the forecasted configuration change for the data center.

Claims

1. A method, comprising:

at a data center controller;

determining a first server location from a plurality of available server locations at which to change a server active state for a data center;

forecasting a time and configuration change for the data center;

determining a second server location from the plurality of available server locations at which to change the server active state based upon the forecasted configuration change for the data center; and

selecting an available server location at which to change the server active state based upon the first server location, the second server location, and the forecasted time associated with the forecasted configuration change for the data center.

2. The method of claim 1:

wherein determining the first server location from the plurality of available server locations at which to change the server active state for the data center comprises;

calculating a first heat potential value for each of the plurality of available server locations based upon currently-active servers; and

selecting a first optimal heat potential value from the calculated first heat potential value for each of the plurality of available server locations; and

wherein determining the second server location from the plurality of available server locations at which to change the server active state based upon the forecasted configuration change for the data center comprises;

calculating a second heat potential value for each of the plurality of available server locations based upon a number of active servers required to service the forecasted configuration change for the data center; and

selecting a second optimal heat potential value from the calculated second heat potential value for each of the plurality of available server locations.

3. The method of claim 2, wherein selecting the available server location at which to change the server active state based upon the first server location, the second server location, and the forecasted time associated with the forecasted configuration change for the data center comprises:

comparing the first optimal heat potential value with the second optimal heat potential value;

calculating an average heat potential value of the first optimal heat potential value and the second optimal heat potential value; and

identifying, from the calculated second heat potential value for each of the plurality of available server locations, a server location with a calculated second heat potential value proximate to the calculated average heat potential value as the available server location at which to change the server active state.

4. The method of claim 1, wherein the forecasted configuration change for the data center comprises one of an increase and a decrease in computational load and wherein determining the second server location from the plurality of available server locations at which to change the server active state based upon the forecasted configuration change for the data center comprises:

upon determining that the forecasted configuration change for the data center comprises the increase in computational load;

calculating a first plurality of heat potential values for the plurality of available server locations based upon a number of active servers required to service the increase in computational load;

determining a minimum heat potential value from the first plurality of heat potential values; and

identifying the second server location as a server location comprising the minimum heat potential value; and

upon determining that the forecasted configuration change for the data center comprises the decrease in computational load;

calculating the first plurality of heat potential values for the plurality of available server locations based upon a number of active servers required to service the decrease in computational load;

determining a maximum heat potential value from the first plurality of heat potential values; and

identifying the second server location as a server location comprising the maximum heat potential value.

5. The method of claim 1, further comprising:

determining whether the forecasted time associated with the forecasted configuration change for the data center exceeds a configured threshold time for changes to selection of servers from the plurality of available server locations; and

wherein, upon determining that the forecasted time associated with the forecasted configuration change for the data center does not exceed the configured threshold time for changes to selection of servers, selecting the available server location at which to change the server active state based upon the first server location, the second server location, and the forecasted time associated with the forecasted configuration change for the data center comprises selecting the first server location at which to change an active state for a server.

6. The method of claim 1, wherein forecasting the configuration change for the data center comprises:

analyzing current load on servers that are active within the data center;

determining an anticipated magnitude of a load change associated with a change in data center computational load; and

determining a change in a number of servers within the data center to accommodate the anticipated magnitude of the load change associated with the change in the data center computational load.

7. The method of claim 1, further comprising:

determining that the forecasted time associated with the forecasted configuration change for the data center exceeds a configured time threshold for server distribution changes;

redistributing active states among servers at the plurality of available server locations to achieve the determined redistribution; and

changing an active state for a server within the redistribution of active servers.

8. A system, comprising:

a memory; and

a data center controller configured to;

determine a first server location from a plurality of available server locations at which to change a server active state for a data center;

forecast a time and configuration change for the data center;

determine a second server location from the plurality of available server locations at which to change the server active state based upon the forecasted configuration change for the data center; and

select an available server location at which to change the server active state based upon the first server location, the second server location, and the forecasted time associated with the forecasted configuration change for the data center.

9. The system of claim 8:

wherein, in being configured to determine the first server location from the plurality of available server locations at which to change the server active state for the data center, the data center controller is configured to;

calculate a first heat potential value for each of the plurality of available server locations based upon currently-active servers; and

select a first optimal heat potential value from the calculated first heat potential value for each of the plurality of available server locations; and

wherein, in being configured to determine the second server location from the plurality of available server locations at which to change the server active state based upon the forecasted configuration change for the data center, the data center controller is configured to;

calculate a second heat potential value for each of the plurality of available server locations based upon a number of active servers required to service the forecasted configuration change for the data center; and

select a second optimal heat potential value from the calculated second heat potential value for each of the plurality of available server locations.

10. The system of claim 9, wherein, in being configured to select the available server location at which to change the server active state based upon the first server location, the second server location, and the forecasted time associated with the forecasted configuration change for the data center, the data center controller is configured to:

compare the first optimal heat potential value with the second optimal heat potential value;

calculate an average heat potential value of the first optimal heat potential value and the second optimal heat potential value; and

identify, from the calculated second heat potential value for each of the plurality of available server locations, a server location with a calculated second heat potential value proximate to the calculated average heat potential value as the available server location at which to change the server active state.

11. The system of claim 8, wherein the forecasted configuration change for the data center comprises one of an increase and a decrease in computational load and wherein, in being configured to determine the second server location from the plurality of available server locations at which to change the server active state based upon the forecasted configuration change for the data center, the data center controller is configured to:

upon determining that the forecasted configuration change for the data center comprises the increase in computational load;

calculate a first plurality of heat potential values for the plurality of available server locations based upon a number of active servers required to service the increase in computational load;

determine a minimum heat potential value from the first plurality of heat potential values; and

identify the second server location as a server location comprising the minimum heat potential value; and

upon determining that the forecasted configuration change for the data center comprises the decrease in computational load;

calculate the first plurality of heat potential values for the plurality of available server locations based upon a number of active servers required to service the decrease in computational load;

determine a maximum heat potential value from the first plurality of heat potential values; and

identify the second server location as a server location comprising the maximum heat potential value.

12. The system of claim 8, wherein the data center controller is further configured to:

determine whether the forecasted time associated with the forecasted configuration change for the data center exceeds a configured threshold time for changes to selection of servers from the plurality of available server locations; and

wherein, upon determining that the forecasted time associated with the forecasted configuration change for the data center does not exceed the configured threshold time for changes to selection of servers, the data center controller is configured to select the first server location at which to change an active state for a server.

13. The system of claim 8, wherein the data center controller is further configured to:

determine that the forecasted time associated with the forecasted configuration change for the data center exceeds a configured time threshold for server distribution changes;

redistribute active states among servers at the plurality of available server locations to achieve the determined redistribution; and

change an active state for a server within the redistribution of active servers.

14. A computer program product comprising:

a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code comprising;

computer readable program code configured to determine a first server location from a plurality of available server locations at which to change a server active state for a data center;

computer readable program code configured to forecast a time and configuration change for the data center;

computer readable program code configured to determine a second server location from the plurality of available server locations at which to change the server active state based upon the forecasted configuration change for the data center; and

computer readable program code configured to select an available server location at which to change the server active state based upon the first server location, the second server location, and the forecasted time associated with the forecasted configuration change for the data center.

15. The computer program product of claim 14:

wherein the computer readable program code configured to determine the first server location from the plurality of available server locations at which to change the server active state for the data center comprises;

computer readable program code configured to calculate a first heat potential value for each of the plurality of available server locations based upon currently-active servers; and

computer readable program code configured to select a first optimal heat potential value from the calculated first heat potential value for each of the plurality of available server locations; and

wherein the computer readable program code configured to determine the second server location from the plurality of available server locations at which to change the server active state based upon the forecasted configuration change for the data center comprises;

computer readable program code configured to calculate a second heat potential value for each of the plurality of available server locations based upon a number of active servers required to service the forecasted configuration change for the data center; and

computer readable program code configured to select a second optimal heat potential value from the calculated second heat potential value for each of the plurality of available server locations.

16. The computer program product of claim 15, wherein the computer readable program code configured to select the available server location at which to change the server active state based upon the first server location, the second server location, and the forecasted time associated with the forecasted configuration change for the data center comprises:

computer readable program code configured to compare the first optimal heat potential value with the second optimal heat potential value;

computer readable program code configured to calculate an average heat potential value of the first optimal heat potential value and the second optimal heat potential value; and

computer readable program code configured to identify, from the calculated second heat potential value for each of the plurality of available server locations, a server location with a calculated second heat potential value proximate to the calculated average heat potential value as the available server location at which to change the server active state.

17. The computer program product of claim 14, wherein the forecasted configuration change for the data center comprises one of an increase and a decrease in computational load, and wherein the computer readable program code configured to determine the second server location from the plurality of available server locations at which to change the server active state based upon the forecasted configuration change for the data center comprises:

upon determining that the forecasted configuration change for the data center comprises the increase in computational load;

computer readable program code configured to calculate a first plurality of heat potential values for the plurality of available server locations based upon a number of active servers required to service the increase in computational load;

computer readable program code configured to determine a minimum heat potential value from the first plurality of heat potential values; and

computer readable program code configured to identify the second server location as a server location comprising the minimum heat potential value; and

upon determining that the forecasted configuration change for the data center comprises the decrease in computational load;

computer readable program code configured to calculate the first plurality of heat potential values for the plurality of available server locations based upon a number of active servers required to service the decrease in computational load;

computer readable program code configured to determine a maximum heat potential value from the first plurality of heat potential values; and

computer readable program code configured to identify the second server location as a server location comprising the maximum heat potential value.

18. The computer program product of claim 14, further comprising:

computer readable program code configured to determine whether the forecasted time associated with the forecasted configuration change for the data center exceeds a configured threshold time for changes to selection of servers from the plurality of available server locations; and

wherein, upon determining that the forecasted time associated with the forecasted configuration change for the data center does not exceed the configured threshold time for changes to selection of servers, the computer readable program code configured to select the available server location at which to change the server active state based upon the first server location, the second server location, and the forecasted time associated with the forecasted configuration change for the data center comprises computer readable program code configured to select the first server location at which to change an active state for a server.

19. The computer program product of claim 14, wherein the computer readable program code configured to forecast the configuration change for the data center computational load comprises:

computer readable program code configured to analyze current load on servers that are active within the data center;

computer readable program code configured to determine an anticipated magnitude of a load change associated with the a change in data center computational load; and

computer readable program code configured to determine a change in a number of servers within the data center to accommodate the anticipated magnitude of the load change associated with the change in the data center computational load.

20. The computer program product of claim 14, further comprising:

computer readable program code configured to determine that the forecasted time associated with the forecasted configuration change for the data center exceeds a configured time threshold for server distribution changes;

computer readable program code configured to redistribute active states among servers at the plurality of available server locations to achieve the determined redistribution; and

computer readable program code configured to change an active state for a server within the redistribution of active servers.