Device, system, and method of classifying a workload of a software service

1. A method of classifying a workload of a software service, the method comprising:

sampling a plurality of values of at least one parameter of the software service by performing out-of-band monitoring of said parameter, wherein changes in said parameter values indicate a change in the workload of the software service in real-time;

determining a workload signature for the software service based on statistical analysis of said sampling of plurality of values, wherein said workload signature provides a representation of the workload of the software service for a first period associated with said sampling;

comparing the workload signature with a plurality of predefined workload signatures stored in a workload signature database, wherein each of the plurality of predefined workload signatures are respectively associated with a corresponding workload classification from among a plurality of predefined workload classifications; and

matching the workload of said software service for the first period with a first workload classification from among the plurality of predefined workload classifications;

wherein the workload of the software service for a second period of real-time sampling is reclassified to a second workload classification, based on a workload signature calculated for the second period of time matching the second workload classification.

2. The method of claim 1 , further comprising:

allocating first resources to the software service according to the first workload classification, and

allocating second resources to the software service according to the second workload classification.

3. The method of claim 2 , wherein determining said workload signature comprises:

determining a plurality of statistical values by applying a statistical function to said plurality of values; and

determining said workload signature based on at least one of said statistical values.

4. The method of claim 3 , wherein said statistical values comprise a plurality of occurrence values, each representing a number of occurrences of a value of said plurality of values, and wherein determining said workload signature comprises determining said workload signature based on a predefined number of maximal values of said plurality of occurrence values.

5. The method of claim 2 , wherein said plurality of workload classifications correspond to a plurality of predefined workload schemes, and wherein each of said plurality of workload signatures comprises one or more statistical values corresponding to one or more values of said parameter when said software service is subject to a workload scheme of said plurality of workload schemes.

6. The method of claim 1 comprising:

sampling another plurality of values of said at least one parameter; and

re-classifying the workload of said software service by re-selecting a workload classification from said plurality of predefined workload classifications based on said other plurality of values.

7. The method of claim 1 , wherein said software service is a virtual machine or an application program.

8. A computing system, the system comprising:

one or more processors for executing a software service, wherein program code for the software service is stored on one or more data storage mediums and is loaded into memory for execution;

an out-of-band monitor for sampling a plurality of values of at least one parameter of the software service by performing out-of-band monitoring of said parameter, wherein changes in said parameter values indicate a change in the workload of the software service in real-time;

a workload classifier for:

determining a workload signature for the software service based on statistical analysis of said sampling of plurality of values, wherein said workload signature provides a representation of the workload of the software service for a first period associated with said sampling;

comparing the workload signature with a plurality of predefined workload signatures stored in a workload signature database, wherein each of the plurality of predefined workload signatures are respectively associated with a corresponding workload classification from among a plurality of predefined workload classifications; and

matching the workload of said software service for the first period with a first workload classification from among the plurality of predefined workload classifications;

wherein the workload of the software service for a second period of real-time sampling is reclassified to a second workload classification, based on a workload signature calculated for the second period of time matching the second workload classification.

9. The computing system of claim 8 , wherein first resources are allocated to the software service according to the first workload classification and second resources are allocated to the software service according to the second workload classification.

10. The computing system of claim 9 , wherein said classifier is capable of determining a plurality of statistical values by applying a statistical function to said plurality of values; and determining said workload signature based on at least one of said statistical values.

11. The computing system of claim 10 , wherein said statistical values comprise a plurality of occurrence values, each representing a number of occurrences of a value of said plurality of values, and wherein said classifier is capable of determining said workload signature based on a predefined number of maximal values of said plurality of occurrence values.

12. The computing system of claim 9 , wherein said plurality of workload classifications correspond to a plurality of predefined workload schemes, and wherein each of said plurality of workload signatures comprises one or more statistical values corresponding to one or more values of said workload-related parameter when said software service is subject to a workload scheme of said plurality of workload schemes.

13. The computing system of claim 8 , wherein said classifier is capable of re-classifying the workload of said software service by re-selecting a workload classification from said plurality of predefined workload classifications based on another plurality of sampled values of said at least one parameter.

14. The computing system of claim 8 , wherein said software service is a virtual machine or an application program.

15. A computer program product comprising a non-transient computer-useable storage medium including a computer-readable program, wherein the computer-readable program when executed on a computer causes the computer to:

sample a plurality of values of at least one parameter of the software service by performing out-of-band monitoring of said parameter, wherein changes in said parameter values indicate a change in the workload of the software service in real-time;

determine a workload signature for the software service based on statistical analysis of said sampling of plurality of values, wherein said workload signature provides a representation of the workload of the software service for a first period associated with said sampling;

compare the workload signature with a plurality of predefined workload signatures stored in a workload signature database, wherein each of the plurality of predefined workload signatures are respectively associated with a corresponding workload classification from among a plurality of predefined workload classifications; and

match the workload of said software service for the first period with a first workload classification from among the plurality of predefined workload classifications,

wherein the workload of the software service for a second period of real-time sampling is reclassified to a second workload classification, based on a workload signature calculated for the second period of time matching the second workload classification.

16. The computer program product of claim 15 , wherein first resources are allocated to the software service according to the first workload classification and second resources are allocated to the software service according to the second workload classification.

17. The computer program product of claim 16 , wherein said computer-readable program causes said computer to determine a plurality of statistical values by applying a statistical function to said plurality of values; and determine said workload signature based on at least one of said statistical values.

18. The computer-readable program of claim 17 , wherein said statistical values comprise a plurality of occurrence values, each representing a number of occurrences of a value of said plurality of values, and wherein said computer-readable program causes said computer to determine said workload signature based on a predefined number of maximal values of said plurality of occurrence values.

19. The computer-readable program of claim 16 , wherein said plurality of workload classifications correspond to a plurality of predefined workload schemes, and wherein each of said plurality of workload signatures comprises one or more statistical values corresponding to one or more values of said workload-related parameter when said software service is subject to a workload scheme of said plurality of workload schemes.

20. The computer-readable program of claim 15 , wherein said computer-readable program causes said computer to sample another plurality of values of said at least one parameter; and re-classify the workload of said software service by re-selecting a workload classification from said plurality of predefined workload classifications based on said other plurality of values.