System and method for selecting and scheduling corrective actions for automated storage management

• US 8,326,669 B2
• Filed: 04/19/2007
• Issued: 12/04/2012
• Est. Priority Date: 04/19/2007
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Korupolu, Madhukar R., Uttamchandani, Sandeep M., Voruganti, Kaladhar, Yin, Li
• Primary Examiner(s): Jarrett, Scott L
• Assistant Examiner(s): CHOY, PAN G
• Application Number:
  US11/737,249
• Publication Number:
  US 20080262890A1
• Time in Patent Office: 2,056 Days
• Field of Search: 705/7.11, 705/7.12, 705/7.22, 705/7.25
• US Class Current: 705/7.22
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/06312 Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
• CPC Class Codes: G06Q 10/06316 Sequencing of tasks or work
• CPC Class Codes: G06Q 10/0635 Risk analysis of enterprise or organisation activities

Abstract

Disclosed is a storage management framework that integrates corrective action plans output from multiple different types of planning tools, sorts the different corrective action plans based on utility and risk values and outputs a time-based schedule for implementing one or more of the corrective action plans to resolve identified current and anticipated workload service level objective (SLO) violations.

Claims

1. A management system for a data storage system comprising:

a controller receiving inputs, said inputs comprising;

business constraints, said business constraints comprising at least service level objectives (SLOs) associated with performance of said data storage system;

system attributes; and

a triggering event comprising an event indicative of at least one of a current violation and an anticipated violation of one of said service level objectives (SLOs); and

at least two different types of planning tools for operating said data storage system, said planning tools being in communication with said controller, and comprising at least two of the following;

a workload throttling planning tool, a buffer size planning tool, a data migration planning tool and a new hardware planning tool; and

a calculator in communication with said controller and said planning tools,

said controller further performing the following;

automatically determining, based on said inputs, whether corrective action is warranted in response to said triggering event;

when said corrective action is warranted in response to said triggering event, generating a plurality of queries comprising different time periods; and

when said corrective action is not warranted in response to said triggering event, taking no further action respect to said triggering event;

said planning tools receiving said queries, each of said planning tools generating and outputting multiple corrective action plans for responding to said one of said current violation and said anticipated violation of said one of said service level objectives (SLOs) such that each corrective action plan is optimized for responding to said triggering event during a corresponding one of said different time periods and such that each corrective action plan specifies parameters for implementing said corrective action plan, and

said calculator calculating risk and utility values for each of said multiple corrective action plans based on said parameters, said business constraints and said system attributes, wherein the utility value of a corrective action plan indicates an expected workload benefit to said data storage system when said corrective action plan is implemented and the risk value of said corrective action plan indicates a probability that said expected workload benefit will be lost as a result of any of volatility in workload time-series functions and human error, and

said controller comparing and ranking, for a specified time period, said risk and utility values of said multiple corrective action plans, selecting at least one of said multiple corrective action plans to perform during said specified time period based on said ranking and creating and outputting a schedule for implementing said at least one of said multiple corrective action plans in said specified time period.

2. The management system of claim 1, further comprising:

a translator in communication with said controller and said planning tools, said translator translating said query into appropriate formats for input into each of said planning tools.

3. The management system of claim 1,

said business constraints comprising utility functions, budget constraints, a desired analysis window and said service level objectives (SLOs), and

said system attributes comprising current and anticipated system states.

4. The management system of claim 1,

said controller further generating, when said corrective action is warranted in response to said triggering event, a plurality of queries specifying different time periods, and

said planning tools each outputting multiple corrective action plans for responding to said one of said current violation and said anticipated violation of said one of said service level objectives (SLOs), each corrective action plan being optimized for responding to said triggering event during a corresponding one of said different time periods.

5. The management system of claim 1, said schedule comprising a time-based schedule comprising:

more than one of said corrective action plans;

a sequence for implementing said corrective action plans; and

approximate times for initiating and completing said corrective action plans.

6. The management system of claim 1,

said schedule further comprising checkpoint states each of which comprises an anticipated state of said data storage system during said implementing of said at least one of said corrective action plans according to said schedule, and

said controller altering said schedule based on differences between an actual state of said data storage system at a given time and a checkpoint state for said given time.

7. A management system for a data storage system comprising:

a controller receiving inputs, said inputs comprising;

business constraints, said business constraints comprising at least service level objectives (SLOs) associated with performance of said data storage system;

system attributes; and

a triggering event comprising an event indicative of at least one of a current violation and an anticipated violation of one of said service level objectives (SLOs);

at least two different types of planning tools for operating said data storage system, said planning tools being in communication with said controller and comprising at least two of the following;

a workload throttling planning tool, a buffer size planning tool, a data migration planning tool and a new hardware planning tool; and

a calculator in communication with said controller and said planning tools,

said controller further performing the following;

automatically determining, based on said inputs, whether corrective action is warranted in response to said triggering event;

when said corrective action is warranted in response to said triggering event, generating a plurality of queries comprising different time periods; and

,

when said corrective action is not warranted in response to said triggering event, taking no further action respect to said triggering event;

said planning tools receiving said queries, each of said planning tools generating and outputting multiple corrective action plans for responding to said one of said current violations and said anticipated violation of one of said service level objectives (SLOs) such that each corrective action plan is optimized for responding to said triggering event during a corresponding one of said different time periods and such that each corrective action plan specifies parameters for implementing said corrective action plan,

said calculator calculating risk and utility values for each of said multiple corrective actions based on said parameters, said business constraints and said system attributes, wherein the utility value of a corrective action plan indicates an expected workload benefit to said data storage system when said corrective action plan is implemented and the risk value of said corrective action plan indicates a probability that said expected workload benefit will be lost as a result of any of volatility in workload time-series functions and human error, and

said controller comparing and ranking, for a specified time period, said risk and utility values of said multiple corrective action plans, selecting multiple ones said multiple corrective action plans to perform during said specified time period based on said ranking, and creating and outputting a schedule for implementing said multiple ones of said multiple corrective action plans in said specified time period.

8. The management system of claim 7,

said business constraints further comprising utility functions, budget constraints, and a desired analysis window, and

said system attributes comprising current and anticipated system states.

9. The management system of claim 7, further comprising a translator in communication with said controller and said planning tools, said translator translating said queries into appropriate formats for input into each of said planning tools.

10. The management system of claim 7, said schedule further comprising a time-based schedule comprising:

a sequence for implementing said selected ones of said multiple corrective action plans; and

approximate times for initiating and completing said selected ones of said multiple corrective action plans.

11. The management system of claim 7,

said schedule further comprising checkpoint states, each of which comprises an anticipated state of said data storage system during said implementing of said selected ones of said multiple corrective action plans according to said schedule, and

aid controller further alters said schedule based on differences between an actual state of said data storage system at a given time and a checkpoint state for said given time.

12. A computer-implemented method for managing a data storage system, said method comprising:

receiving, by a computer, inputs, said inputs comprising;

business constraints, said business constraints comprising at least service level objectives (SLOs) associated with performance of said data storage system;

system attributes; and

a triggering event comprising an event indicative of at least one of a current violation and an anticipated violation of one of said service level objectives (SLOs);

automatically determining, by said computer and based on said inputs, whether corrective action is warranted in response to said triggering event;

when said corrective action is warranted in response to said triggering event, generating, by said computer, plurality of queries comprising different time periods;

when said correction action is not warranted in response to said triggering event, taking, by said computer, no further action in response to said triggering event;

receiving, by at least two different types of planning tools of said computer for operating said data storage system, said queries, each of said planning tools further generating and outputting multiple corrective action plans for responding to said one of said current violation and said anticipated violation of said one of said service level objectives (SLOs) such that each corrective action plan is optimized for responding to said triggering event during a corresponding one of said different time periods and such that each corrective action plan specifies parameters for implementing said corrective action plan, said planning tools comprising at least two of the following;

a workload throttling planning tool, a buffer size planning tool, a data migration planning tool and a new hardware planning tool;

calculating, by said computer, risk and utility values for each of said multiple corrective action plans based on said parameters, said business constraints and said system attributes, wherein the utility value of a corrective action plan indicates an expected workload benefit to said data storage system when said corrective action plan is implemented and the risk value of said corrective action plan indicates a probability that said expected workload benefit will be lost as a result of any of volatility in workload time-series functions and human error;

comparing and ranking, by said computer and for a specified time period, said risk and utility values of said multiple corrective action plans and selecting at least one of said multiple corrective actions to perform during said specified time period based on said ranking, and

automatically creating and outputting, by said computer, a schedule for implementing said at least one of said multiple corrective action plans in said specified time period.

13. The method of claim 12,

said business constraints comprising utility functions, budget constraints, a desired analysis window and said service level objectives (SLOs), said service level objectives (SLOs) comprising service-level objective (SLO) violation thresholds, and

said system attributes comprising current and anticipated system states.

14. The method of claim 12,

said generating further comprising generating, when said corrective action is warranted in response to said triggering event, a plurality of queries specifying different time periods, and

said creating and outputting further comprising creating and outputting multiple corrective action plans for responding to said one of said current violation and said anticipated violation of said one of said service level objectives (SLOs), each corrective action plan being optimized to respond to said triggering event during a corresponding one of said different time periods.

15. The method of claim 12, said creating of said schedule further comprising:

selecting more than one of said corrective action plans;

determining a sequence for said implementing of said more than one of said corrective action plans; and

determining approximate times for initiating said implementing of said more than one of said corrective action plans.

16. The method of claim 12, further comprising continuously monitoring an impact of any corrective action plan implemented in said data storage system.

17. The method of claim 12, further comprising automatically translating, by said computer system said query into appropriate formats for input into each of said planning tools.

18. A program storage device readable by computer and tangibly embodying a program of instructions that, when executed by said computer, perform a method of managing a data storage system, said method comprising:

receiving inputs, said inputs comprising;

business constraints, said business constraints comprising at least service level objectives (SLOs) associated with performance of said data storage system;

system attributes; and

a triggering event comprising an event indicative of at least one of a current violation and an anticipated violation of one of said service level objectives (SLOs);

automatically determining, based on said inputs, whether corrective action is warranted in response to a triggering event;

when said corrective action is warranted in response to said triggering event, generating a plurality of queries comprising different time periods;

when said correction action is not warranted in response to said triggering event, taking no further action in response to said triggering event;

receiving, by at least two different planning tools for operating said data storage system, said queries, each of said planning tools further generating and outputting multiple corrective action plans for responding to said one of said current violation and said anticipated violation of said one of said service level objectives (SLOs) such that each corrective action plan is optimized for responding to said triggering event during a corresponding one of said different time periods and such that each corrective action plan specifies parameters for implementing said corrective action plan, said planning tools comprising at least two of the following;

a workload throttling planning tool, a buffer size planning tool, a data migration planning tool and a new hardware planning tool;

calculating risk and utility values for each of said multiple corrective action plans based on said parameters, said business constraints and said system attributes, wherein the utility value of a corrective action plan indicates an expected workload benefit to said data storage system when said corrective action plan is implemented and the risk value of said corrective action plan indicates a probability that said expected workload benefit will be lost as a result of any of volatility in workload time-series functions and human error;

comparing and ranking, for a specified time period, said risk and utility values of said multiple corrective action plans and selecting at least one of said multiple corrective actions to perform during said specified time period based on ranking; and

automatically creating and outputting a schedule for implementing said at least one of said multiple corrective action plans in said specified time period.

19. The program storage device of claim 18,

said business constraints comprising utility functions, budget constraints, a desired analysis window and said service level objectives (SLOs), said service level objectives (SLOs) comprising service-level objective violation thresholds, and

said system attributes comprising current and anticipated system states.

20. The program storage device of claim 18,

said generating further comprising generating, when said corrective action is warranted in response to said triggering event, a plurality of queries specifying different time periods so that multiple corrective action plans, and

said creating and outputting further comprising creating and outputting multiple corrective action plans for responding to said one of said current violation and said anticipated violation of said one of said service level objectives (SLOs), each corrective action plan being optimized to respond to said triggering event during a corresponding one of said different time periods.

21. The program storage device of claim 18, said creating of said schedule further comprising:

selecting more than one of said corrective action plans;

determining a sequence for said implementing of said more than one of said corrective action plans; and

determining approximate times for initiating said implementing of said more than one of said corrective action plans.

22. The program storage device of claim 18, said method further comprising continuously monitoring an impact of any corrective action plan implemented in said data storage system.

23. The program storage device of claim 18, said method further comprising automatically translating said query into appropriate formats for input into each of said planning tools.