Reservation of hardware resources in a computer system based on utilization measurements during time ranges

• US 10,120,718 B2
• Filed: 08/24/2016
• Issued: 11/06/2018
• Est. Priority Date: 08/24/2016
• Status: Active Grant

• Current Assignee: CA, Inc. (d/b/a CA Technologies) (Broadcom, Inc.)
• Original Assignee: CA, Inc. (d/b/a CA Technologies) (Broadcom, Inc.)
• Inventors: Esterkin, Alex, Rich, David, Mercer, James
• Primary Examiner(s): Zand, Davoud A
• Application Number:
  US15/245,907
• Publication Number:
  US 20180060127A1
• Time in Patent Office: 804 Days
• CPC Class Codes: G06F 1/3287 by switching off individual functional units in the computer system
• CPC Class Codes: G06F 11/3006 where the computing system is distributed, e.g. networked systems, clusters, multiprocessor systems multiprogramming arrangements G06F9/46; allocation of resources G06F9/50
• CPC Class Codes: G06F 11/302 where the computing system component is a software system
• CPC Class Codes: G06F 11/3055 Monitoring arrangements for monitoring the status of the computing system or of the computing system component, e.g. monitoring if the computing system is on, off, available, not available error or fault processing without redundancy G06F11/0703; error detection or correction by redundancy in data representation G06F11/08; error detection or correction by redundancy in operation G06F11/14; error detection or correction by redundancy in hardware G06F11/16
• CPC Class Codes: G06F 11/327 Alarm or error message display
• CPC Class Codes: G06F 11/3409 for performance assessment
• CPC Class Codes: G06F 11/3452 Performance evaluation by statistical analysis
• CPC Class Codes: G06F 2201/81 Threshold
• CPC Class Codes: G06F 2201/88 Monitoring involving counting
• CPC Class Codes: G06F 2209/5014 Reservation
• CPC Class Codes: G06F 9/5011 the resources being hardware resources other than CPUs, Servers and Terminals
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources

Abstract

A resource management computer node obtains hardware utilization values measured for a hardware resource of a computer system being used by a software application. For a set of the utilization values that were measured during a same time-of-day range on a same day of week, the node determines a count value indicating a number of times the utilization values in the set exceed a count threshold, determines a count percentage based on a ratio of the count value to a sum of count values determined for the same day of week, compares the count percentage to a busy threshold, and, responsive to if the count percentage is determined to exceed the busy threshold, sets a busy indicator object at a location in a resource utilization data structure having a defined correspondence to the time-of-day range. The node controls reservation of hardware resources for the software application responsive to whether the busy indicator object has been set.

Claims

1. A method performed by a resource management computer node comprising:

obtaining hardware utilization values measured for a hardware resource of a computer system being used by a software application;

for a set of the hardware utilization values that were measured during a same time-of-day range on a same day of week,

determining a count value indicating a number of times the hardware utilization values in the set exceed a count threshold value,

determining a count percentage value based on a ratio of the count value to a sum of count values determined for the same day of week,

comparing the count percentage value to a busy threshold value, and

responsive to if the count percentage value is determined to exceed the busy threshold value,

setting a busy indicator object at a location in a resource utilization data structure having a defined correspondence to the time-of-day range;

controlling reservation of hardware resources for operational use by the software application during at least the time-of-day range responsive to whether the busy indicator object has been set at the location in the resource utilization data structure; and

subsequent to the controlling reservation of hardware resources, running the software application on the reserved hardware resources during at least the time-of-day range.

2. The method of claim 1, wherein:

the controlling reservation of hardware resources comprises,

reserving a defined plural number of computer servers for use by the software application during at least the time-of-day range while precluding other software applications from using the defined plural number of computer servers that are reserved responsive to determining that the busy indicator object has been set at the location in the resource utilization data structure; and

the running the software application on the reserved hardware resources during at least the time-of-day range comprises,

running the software application on the defined plural number of computer servers that are reserved during at least the time-of-day range while precluding other software applications from using the defined plural number of computer servers.

3. The method of claim 1, wherein:

the hardware utilization values are measured as at least one of processor utilization, non-volatile memory device storage space utilization, and data network communication utilization.

4. The method of claim 1, further comprising:

determining the count threshold value based on a numerical combination of the hardware utilization values in the set.

5. The method of claim 4, wherein the count threshold value is determined based on a numerical average of the hardware utilization values in the set.

6. The method of claim 1, further comprising:

repeating for each of a series of time-of-day ranges during each day of a defined plurality of weeks;

the determining a count value;

the determining a count percentage value;

the comparing the count percentage value to a busy threshold value; and

the responsive to if the count percentage value is determined to exceed the busy threshold value, setting a busy indicator object at a location in a resource utilization data structure having a defined correspondence to the time-of-day range; and

wherein the reservation of hardware resources for operational use by the software application is controlled based on whether the busy indicator objects have been set at locations in the resource utilization data structure for at least a threshold number of the time-of-day ranges.

7. The method of claim 6, wherein the reservation of hardware resources for operational use by the software application is controlled based on whether the busy indicator objects have been set at locations in the resource utilization data structure for at least the threshold number of the time-of-day ranges within at least a threshold number of the days of the plurality of weeks.

8. The method of claim 6, further comprising:

determining the busy threshold value based on a numerical combination of the count percentage values determined for each of the series of time-of-day ranges during each day of the plurality of weeks.

9. The method of claim 1, further comprising:

obtaining hardware utilization values measured for a plurality of other hardware resources of the computer system being used by the software application;

repeating for each of the other hardware resources and for another set of the hardware utilization values that were measured for the other hardware resource during the same time-of-day range on the same day of week,

the determining a count value indicating a number of times the hardware utilization values in the another set exceed the count threshold value,

the determining a count percentage value based on a ratio of the count value to a sum of count values determined for the same day of week,

the comparing the count percentage value to the busy threshold value, and

the responsive to if the count percentage value is determined to exceed the busy threshold value,

setting a busy indicator object at another location in the resource utilization data structure having a defined correspondence to the time-of-day range and the other hardware resource; and

controlling reservation of hardware resources for operational use by the software application during at least the time-of-day range responsive to whether the busy indicator objects have been set at the locations in the resource utilization data structure.

10. The method of claim 1, further comprising:

determining an elapsed time between initiation of start-up of the hardware resource from a low-power or power-off state and for the hardware resource to reach an operational mode that allows operational use by the software application; and

determining a weighting factor based on comparison of the start-up time to a threshold time,

wherein the count percentage value is determined based on combining the weighting factor with the ratio of the count value to the sum of count values determined for the same day of week.

11. The method of claim 1, further comprising:

determining an error rate for operation of the hardware resource; and

determining a weighting factor based on comparison of the error rate to a threshold rate,

wherein the count percentage value is determined based on combining the weighting factor with the ratio of the count value to the sum of count values determined for the same day of week.

12. The method of claim 1, further comprising:

determining an estimate of remaining operational life of the hardware resource before failure; and

determining a weighting factor based on comparison of the estimate of remaining operational life to a threshold time duration,

wherein the count percentage value is determined based on combining the weighting factor with the ratio of the count value to the sum of count values determined for the same day of week.

13. The method of claim 1, further comprising:

determining an elapsed time between initiation of start-up of the hardware resource from a low-power or power-off state and for the hardware resource to reach an operational mode that allows operational use by the software application;

decreasing the busy threshold value if the elapsed time exceeds a threshold time; and

increasing the busy threshold value if the elapsed time does not exceed the threshold time.

14. The method of claim 1, further comprising:

determining an error rate for operation of the hardware resource; and

decreasing the busy threshold value if the error rate exceeds a threshold rate; and

increasing the busy threshold value if the error rate does not exceed the threshold rate.

15. The method of claim 1, further comprising:

determining an estimate of remaining operational life of the hardware resource before failure;

decreasing the busy threshold value if the estimate of remaining operational life does not exceed a threshold time duration; and

increasing the busy threshold value if the estimate of remaining operational life exceeds the threshold time duration.

16. The method of claim 1, further comprising:

determining a response latency experienced by the software application performing a call to the hardware resource during the same time-of-day range;

obtaining a maximum response latency constraint defined for the software application; and

determining a weighting factor based on comparison of the input data response latency to the maximum response latency constraint,

wherein the count percentage value is determined based on combining the weighting factor with the ratio of the count value to the sum of count values determined for the same day of week.

17. A resource management computer node comprising:

a network interface circuit configured to communicate with hardware resources of a computer system;

a processor circuit coupled to the network interface circuit; and

a memory circuit coupled to the processor circuit and storing program code which when executed by the processor circuit causes the processor circuit to perform operations comprising;

obtain hardware utilization values measured for at least one of the hardware resources being used by a software application;

for a set of the hardware utilization values that were measured during a same time-of-day range on a same day of week,

determining a count value indicating a number of times the hardware utilization values in the set exceed a count threshold value,

determining a count percentage value based on a ratio of the count value to a sum of count values determined for the same day of week,

comparing the count percentage value to a busy threshold value,

responsive to if the count percentage value is determined to exceed the busy threshold value,

setting a busy indicator object at a location in a resource utilization data structure having a defined correspondence to the time-of-day range; and

controlling reservation of the hardware resources for operational use by the software application during at least the time-of-day range responsive to whether the busy indicator object has been set at the location in the resource utilization data structure; and

subsequent to the controlling reservation of hardware resources, running the software application on the reserved hardware resources during at least the time-of-day range.

18. The resource management computer node of claim 17, wherein the operations further comprise:

obtain hardware utilization values measured for a plurality the hardware resources of the computer system being used by the software application;

repeating for each of the hardware resources and a set of the hardware utilization values that were measured for the hardware resource during the same time-of-day range on the same day of week,

the determining a count value indicating a number of times the hardware utilization values in the set exceed the count threshold value, wherein the count threshold value is determined based on a numerical average of the hardware utilization values in the set,

the determining a count percentage value based on a ratio of the count value to a sum of count values determined for the same day of week,

the comparing the count percentage value to the busy threshold value, and

the responsive to if the count percentage value is determined to exceed the busy threshold value,

setting a busy indicator object at a location in the resource utilization data structure having a defined correspondence to the time-of-day range and the hardware resource; and

controlling reservation of hardware resources for operational use by the software application during at least the time-of-day range responsive to whether the busy indicator objects have been set at the locations in the resource utilization data structure for at least a threshold number of the time-of-day ranges within at least a threshold number of the days of the week,

wherein the busy threshold value is determined based on a numerical average of the count percentage values determined for each of the series of time-of-day ranges during each day of the week.

19. The resource management computer node of claim 17, wherein:

the operations for controlling reservation of hardware resources comprise,

reserving a defined plural number of computer servers for use by the software application during at least the time-of-day range while precluding other software applications from using the defined plural number of computer servers that are reserved responsive to determining that the busy indicator object has been set at the location in the resource utilization data structure; and

the operations for running the software application on the reserved hardware resources during at least the time-of-day range comprise,

running the software application on the defined plural number of computer servers that are reserved during at least the time-of-day range while precluding other software applications from using the defined plural number of computer servers.