Method of controlling access to computing resource within shared computing environment

• US 7,694,305 B2
• Filed: 10/21/2004
• Issued: 04/06/2010
• Est. Priority Date: 10/21/2004
• Status: Expired due to Fees

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Karlsson, Magnus, Zhu, Xiaoyun, Karamanolis, Christos
• Primary Examiner(s): An; Meng-Ai
• Assistant Examiner(s): Truong; Camquy
• Application Number:
  US10/972,232
• Publication Number:
  US 20060090163A1
• Time in Patent Office: 1,993 Days
• Field of Search: 718/105, 709/224, 709/226
• US Class Current: 718/105
• CPC Class Codes: G06F 11/3409 for performance assessment
• CPC Class Codes: G06F 11/3447 Performance evaluation by modeling
• CPC Class Codes: G06F 11/3485 for I/O devices
• CPC Class Codes: G06F 11/349 for interfaces, buses
• CPC Class Codes: G06F 2201/885 Monitoring specific for caches
• CPC Class Codes: G06F 2206/1012 Load balancing
• CPC Class Codes: G06F 3/0613 in relation to throughput
• CPC Class Codes: G06F 3/0635 by changing the path, e.g. traffic rerouting, path reconfiguration
• CPC Class Codes: G06F 3/0653 Monitoring storage devices or systems
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]

Abstract

An embodiment of a method of controlling access to a computing resource within a shared computing environment begins with a first step of measuring performance parameters for workloads accessing the computing resource to determine a performance parameter vector for the workloads. The method continues with a second step of estimating a controller function for the computing resource by analysis of recent performance parameters and recent throughputs. The controller function comprises a mathematical operation which takes an input and provides an output. In a third step, the method compares the performance parameter vector to a reference performance parameter vector to determine an error parameter. In a fourth step, the method applies the controller function to the error parameter to determine a target throughput for each of the workloads. The method concludes with a fifth step of adjusting access to the computing resource for each work load having a throughput limit different from about the target throughput by reducing or increasing the throughput limit for the workload to about the target throughput.

Claims

1. A method of controlling access to a computing resource within a shared computing environment comprising:

measuring performance parameters for workloads accessing the computing resource to determine a performance parameter vector for the workloads;

estimating, by at least one processor, a controller function for the computing resource by analysis of the performance parameters and throughputs associated with accessing the computing resource, the controller function comprising a mathematical operation which takes an input and provides an output;

comparing, by the at least one processor, the performance parameter vector to a reference performance parameter vector to determine an error parameter, wherein the error parameter is determined by subtracting the performance parameter vector from the reference performance parameter vector which forms an error vector and selecting a value from a plurality of values in the error vector as the error parameter;

applying, by the at least one processor, the controller function to the error parameter to determine a target throughput for each of the workloads; and

for each workload having a throughput limit different from the target throughput, adjusting access to the computing resource for the workload by reducing or increasing the throughput limit for the workload according to the target throughput.

2. The method of claim 1 wherein estimating the controller function comprises performing a regression analysis on recent cycles of measured performance parameters and the throughputs.

3. The method of claim 1 wherein the measured performance parameters comprise latencies for clients accessing one or more storage servers.

4. The method of claim 3 wherein the latencies comprise input/output latencies for the clients performing input/output operations on the one more storage servers.

5. The method of claim 3 wherein the throughputs comprise throughputs of input/output operations on the one or more storage servers.

6. The method of claim 1 wherein the workloads comprise desired input/output operations.

7. The method of claim 1 wherein the controller function comprises a linear controller function.

8. The method of claim 1 wherein the performance parameters comprise latencies.

9. The method of claim 1 wherein applying the controller function to the error parameter further comprises applying the controller function to a current aggregate target throughput.

10. The method of claim 1 wherein adjusting the access to the computing resource comprises adjusting a throttling module to reduce or increase the throughput limit.

11. The method of claim 1 further comprising iteratively performing the steps of estimating the controller function, measuring the performance parameters, comparing the performance parameter vector to the reference performance parameter vector, applying the controller function, and adjusting access to the computing resource.

12. The method of claim 1 wherein estimating the controller function further comprises determining whether there have been a sufficient number of requests to make re-estimating the controller function statistically significant.

13. The method of claim 1 wherein estimating the controller function comprises estimating a current model determined in a current cycle and a previous model determined in a previous cycle.

14. The method of claim 13 wherein estimating the controller function determines whether the current model differs substantially from the previous model and, if so, employing the current model for the controller function.

15. The method of claim 13 wherein estimating the controller function determines whether the current model provides a realistic expectation of improved control and, if not, discards the current model and employs the previous model for the controller function.

16. The method of claim 1 wherein estimating the controller function determines whether the controller function provides a realistic operating point and, if not, sets an operating point to a realistic value.

17. The method of claim 1 wherein adjusting the access to the computing resource for the workload employs a throughput allocation table which specifies shares for the workloads according to throughput bands.

18. The method of claim 1 wherein the computing resource is selected from the group consisting of storage, processor cycles, processing units, network bandwidth, memory, servers, another computing resource, and a combination thereof.

19. A method of controlling access to a storage system comprising:

measuring latencies for workloads accessing the storage system to determine a latency vector for the workloads;

estimating, by at least one processor, a controller function for the storage system by analysis of the latencies and throughputs associated with accessing the storage system, the controller function comprising a mathematical operation which takes an input and provides an output;

comparing, by the at least one processor, the latency vector to a reference latency vector to determine an error parameter, wherein the error parameter is determined by subtracting the latency vector from the reference latency vector which forms an error vector and selecting a value from a plurality of values in the error vector as the error parameter;

applying, by the at least one processor, the controller function to the error parameter to determine a target throughput for each of the workloads; and

for each workload having a throughput limit different from the target throughput, adjusting access to the storage system for the workload by reducing or increasing the throughput limit for the workload according to the target throughput.

20. The method of claim 19 wherein estimating the controller function comprises performing a regression analysis on recent cycles of measured latencies and the throughputs.

21. The method of claim 19 wherein the throughputs comprise throughputs of input/output operations on one or more storage servers of the storage system.

22. The method of claim 19 wherein the latencies comprise input/output latencies for clients performing input/output operations on one or more storage servers of the storage system.

23. The method of claim 19 wherein the controller function comprises a linear controller function.

24. The method of claim 19 wherein applying the controller function to the error parameter further comprises applying the controller function to a current aggregate target throughput.

25. The method of claim 19 further comprising iteratively performing the steps of estimating the controller function, measuring the latencies, comparing the latency vector to the reference latency vector, applying the controller function, and adjusting access to the storage system.

26. The method of claim 19 wherein adjusting the access to the storage system for the workload employs a throughput allocation table which specifies shares for the workloads according to throughput bands.

27. A computer readable media comprising computer code that upon execution by at least one processor implements a method of controlling access to a computing resource within a shared computing environment, the method of controlling access to the computing resource comprising:

measuring performance parameters for workloads accessing the computing resource to determine a performance parameter vector for the workloads;

estimating a controller function for the computing resource by analysis of the performance parameters and throughputs associated with accessing the computing resource, the controller function comprising a mathematical operation which takes an input and provides an output;

comparing the performance parameter vector to a reference performance parameter vector to determine an error parameter, wherein the error parameter is determined by subtracting the performance parameter vector from the reference performance parameter vector which forms an error vector and selecting a value from a plurality of values in the error vector as the error parameter;

applying the controller function to the error parameter to determine a target throughput for each of the workloads; and

for each workload having a throughput limit different from the target throughput, adjusting access to the computing resource for the workload by reducing or increasing the throughput limit for the workload to the target throughput.

28. A computer readable media comprising computer code that upon execution by at least one processor implements a method of controlling access to a computing resource within a shared computing environment, the method of controlling access to the computing resource comprising:

measuring latencies for workloads accessing the storage system to determine a latency vector for the workloads;

estimating a controller function for the storage system by analysis of the latencies and throughputs associated with accessing the storage system, the controller function comprising a mathematical operation which takes an input and provides an output;

comparing the latency vector to a reference latency vector to determine an error parameter, wherein the error parameter is determined by subtracting the latency vector from the reference latency vector which forms an error vector and selecting a value from a plurality of values in the error vector as the error parameter;

applying the controller function to the error parameter to determine a target throughput for each of the workloads; and

for each workload having a throughput limit different from the target throughput, adjusting access to the storage system for the workload by reducing or increasing the throughput limit for the workload to the target throughput.