Workload management for maintaining redundancy of non-data computer components

1. A computer system comprising:

data and non-data handling hardware resources, said non-data-handling resources being configured redundantly so that if one fails a workload running on said hardware resources is not interrupted; and

a workload manager for allocating said data hardware resources to workloads in part as a function of at least one of utilization of at least some of said data-handling hardware resources, workload performance, and workload load, and in part as a function of a redundancy level for said non-data handling resources.

2. A computer system as recited in claim 1 wherein; said workload manager, in the event of a failure of a non-data-handling component impairing redundancy so as to impact reliability, reallocates said hardware resources so as to reduce full-utilization performance and to restore redundancy.

3. A computer system as recited in claim 2 wherein said performance is reduced by inactivating some of said hardware resources.

4. A computer system as recited in claim 2 wherein said performance is reduced at least in part by reducing the performance level of at least some of said hardware resources.

5. A computer system as recited in claim 1 wherein said workload manager, in the event of a failure of a non-data-handling component that impairs the reliability of a first set of said hardware resources but not of a second set of said hardware resources, migrates a first workload from said second set to said first set.

6. A computer system as recited in claim 5 wherein said workload manager, in the event of said failure, migrates a second workload from said first set to said second set.

7. A computer system as recited in claim 1 wherein said data-handling components include at least one of processors, storage devices, and communications devices, and wherein said non-data-handling components include at least one of cooling devices and power supplies.

8. A method comprising:

detecting failure of one of plural redundant non-data-handling computer components, said failure resulting in a decrease in the redundancy level of said non-data-handling components;

reducing the overall performance of data-handling computer components formerly serviced by said redundant non-data-handling computer components so as to increase said redundancy level; and

migrating a first workload to said data-handling computer components formerly serviced by said redundant non-data-handling computer components from data-handling computer components not formerly serviced by said redundant non-data-handling computer components.

9. A method as recited in claim 8 wherein said migrating is performed before said reducing.

10. A method as recited in claim 8 further comprising migrating a second workload from said data-handling computer components formerly serviced by said redundant non-data-handling computer components to data-handling computer components not formerly serviced by said redundant non-data-handling computer components.

11. A method as recited in claim 8 wherein said reducing involves inactivating at least one data-handling computer component.

12. A method as recited in claim 8 wherein said reducing involves reducing the clock rate of at least one processor.

13. Computer-readable storage media comprising:

computer-executable software encoded therein, said software including

a data compiler for receiving data indicating that one of plural redundant non-data-handling computer components has failed so as to decrease an associated redundancy level without reducing performance of a workload running on data-handling resources serviced by said non-data-handling computer components;

management policies for indicating how to reallocate hardware resources in response to such a failed one of said plural redundant non-data-handling computer components, said management policies indicating plural factors for planning how to reallocate, said factors including at least one of utilization of said hardware resources, performance of a workload running on said hardware resources, and load on a said workload, said plural factors further including redundancy level of said non-data-handling computer components; and

a workload manager for implementing said reallocation.

14. Computer-readable storage media as recited in claim 13 wherein said policies provide for reducing a performance level of hardware resources serviced by said non-data-handling components to increase said redundancy level.

15. Computer-readable storage media as recited in claim 13 wherein said policies provide for inactivating data-handling resources to increase said redundancy level.

16. Computer-readable storage media as recited in claim 13 wherein said management policies provide for migrating a workload from data-handling resources serviced by said redundant non-data-handling components to other data handling resources not serviced by said non-data-handling components.

17. Computer-readable storage media as recited in claim 13 wherein said management policies provide for migrating a workload to data-handling resources serviced by said redundant non-data-handling components from other data-handling resources not serviced by said non-data-handling components.

18. Computer-readable storage media as recited in claim 13 wherein said non-data-handling components include at least one of power supplies and cooling components and said data-handling resources include at least one of processors, data storage devices, and communications devices.