Early IO control techniques for a remote data facility

Patent metadata

Field	Value
Publication number	US-10846015-B2
Application number	US-201916384859-A
Country	US
Kind code	B2
Filing date	Apr 15, 2019
Priority date	Apr 15, 2019
Publication date	Nov 24, 2020
Grant date	Nov 24, 2020

Abstract

Architectures and techniques are described that can address challenges associated with engaging a procedure that has the potential to produce beneficial effects in some case and detrimental effects in other cases. One example can be engaging an early IO (EIO) protocol in a synchronous remote data facility (SRDF), but only when doing so will have a positive impact on performance and/or response times.

First claim

What is claimed is:

1. A system, comprising: a processor; and
a non-transitory memory that stores executable instructions that, when executed by the processor, facilitate performance of operations, comprising:
determining an action selected from an action group comprising: a first action in which an early I/O procedure is activated and a second action in which the early I/O procedure is not activated;
in response to comparing a current response time of a current I/O operation to a previous response time of a previous I/O operation, determining state data indicative of a current state of the current I/O operation, wherein the current state is selected from a state group comprising a first state indicative of the current response time being less than or equal to the previous response time, and a second state indicative of the current response time being greater than the previous response time; and
determining policy data that identifies a next action from among the action group that is to be applied to a next I/O procedure, wherein the next action is determined, as a function of the current state, to minimize a next response time of the next I/O operation.

2. The system of claim 1 , wherein the operations further comprise instructing a device that manages the early IO procedure to operate according to the policy data.

3. The system of claim 1 , wherein the current IO operation is a write operation that writes data received from a host to a first storage device and a second storage device.

4. The system of claim 1 , wherein the current response time is a first average response time of multiple IO operations performed during a current time window having a fixed duration, the previous response time is a second average response time of multiple IO operations performed during a previous time window having the fixed duration, and the next response time is a third average response time of multiple IO operations performed during a next time window having the fixed duration.

5. The system of claim 4 , wherein the fixed duration is between 100 milliseconds and 500 milliseconds.

6. The system of claim 1 , wherein the determining the state data further comprises determining matrix data, the matrix data comprising:
probability matrixes comprising: a first probability matrix indicative of first probabilities that a state of the state group will occur in response to the first action, and a second probability matrix indicative of second probabilities that the state will occur in response to the second action; and
reward matrixes comprising: a first reward matrix indicative of first reward values assigned to an occurrence of the state in connection with the first probability matrix, and a second reward matrix indicative of second reward values assigned to the occurrence of the state in connection with the second probability matrix.

7. The system of claim 6 , wherein the next action is determined to substantially minimize the next response time based on a determination that the next action has a highest reward value from among the reward values.

8. The system of claim 6 , wherein the determining the policy data comprises performing a regression procedure that iteratively simulates actions of the action group when applied to an IO procedure in states of the state group in order to identify the next action.

9. The system of claim 8 , wherein the regression procedure further comprises a discount factor that reduces a weight of older reward values of the reward values.

10. The system of claim 8 , wherein the regression procedure further comprises a termination factor that facilitates termination of the regression procedure in response to a first iterative result of the regression procedure varying from a second iterative result of the regression procedure by less than the termination factor.

11. A method, comprising:
determining, by a device comprising a processor, an action selected from an action group that comprises a first action in which an early IO procedure is activated and a second action in which the early IO procedure is not activated;
in response to comparing a current response time of a current IO operation to a previous response time of a previous IO operation, determining, by the device, state data indicative of a current state of the current IO operation, wherein the current state is selected from a state group comprising a first state indicative of the current response time being less than or equal to the previous response time, and a second state indicative of the current response time being greater than the previous response time; and
determining, by the device, a policy that identifies a next action from among the action group that is to be applied to a next IO procedure, wherein the next action is determined, as a function of the current state, to substantially minimize a next response time of the next IO operation.

12. The method of claim 11 , further comprising:
performing, by the device, a regression procedure that iteratively simulates actions of the action group when applied to an IO procedure in states of the state group in order to identify the next action.

13. The method of claim 12 , further comprising:
identifying, by the device, the next action in response to determining a highest reward value from among reward values that are assigned to an occurrence of members of the state group.

14. The method of claim 12 , further comprising:
terminating, by the device, the regression procedure in response to a first iterative result of the regression procedure varying from a second iterative result of the regression procedure by less than a defined amount.

Assignees

• Emc Ip Holding Co Llc

Inventors

• Bora Bhaskar
• Goodwill Richard C
• Tannous Toufic

Classifications

• G06F3/0613

  in relation to throughput · CPC title

• G06F3/0611 Primary

  in relation to response time · CPC title

• G06F3/067

  Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS] · CPC title

• G06F3/0659

  Command handling arrangements, e.g. command buffers, queues, command scheduling · CPC title

• G06F3/0655 Primary

  Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices · CPC title