Prescriptive analytics based activation timetable stack for cloud computing resource scheduling

Patent metadata

Field	Value
Publication number	US-10445209-B2
Application number	US-201715811339-A
Country	US
Kind code	B2
Filing date	Nov 13, 2017
Priority date	Sep 8, 2017
Publication date	Oct 15, 2019
Grant date	Oct 15, 2019

Abstract

A multi-layer activation timetable stack may generate prescriptive activation timetables for controlling activation states for computing resources. An input layer of the activation timetable stack may generate time-scaled pattern data. A transformation layer may identify trends and variables at era timescales. A data treatment layer may flag activation states based on the trends identified at the era timescales. Once the activation states a flagged, the prescriptive engine layer may generate an activation timetable that may be used to control computing resource activation prescriptively.

First claim

What is claimed is:

1. A system comprising:
network interface circuitry configured to:
receive historical utilization data for a set of virtual machines;
receive tagging data defining a functional grouping for a first virtual machine of the set of virtual machines;
receive consumption metric data including cost metrics for the set of virtual machines;
receive a threshold selection value for the functional grouping, the threshold selection value including a target efficiency differential that identifies a difference in ratios of utilization time to non-utilization time for the set of virtual machines and further weights the set of virtual machines based on the consumption metric data; and
send an activation timetable to host interface circuitry configured to control activation states for at least the first virtual machine; and

activation timetable circuitry in data communication with the network interface circuitry, the activation timetable circuitry configured to execute an activation timetable stack,
the activation timetable stack comprising:
a data staging layer;
an input layer;
a configuration layer;
a transformation layer;
a data treatment layer; and
a prescriptive engine layer;

the activation timetable circuitry configured to:
parse, at the input layer, the historical utilization data, the consumption metric data, and tagging data to generate time-scaled pattern data;
store, via a database operation at the data staging layer, the time-scaled pattern data;
access, at the transformation layer, the time-scaled pattern data using a memory resource provided by the data staging layer;
identify, at the transformation layer, an era variable indicative of utilization on an era timescale;
set, at the configuration layer, an activation threshold for the first virtual machine responsive to the threshold selection value;
pass the era variable from the transformation layer to the data treatment layer;
responsive to the time-scaled pattern data determine an associated value for an era variable for an activation timeslot;
flag, at the prescriptive engine layer, the activation states based at least in part on the associated value and the activation threshold;
after flagging the activation states, alter an activation state for the activation timeslot responsive to an adjacent activation state for an adjacent timeslot and an intelligent bucket analysis;
generate, at the prescriptive engine layer, the activation timetable for the set of virtual machines, the activation timetable defining activation states for the first virtual machine; and
initiate deployment of the activation timetable by causing the network interface circuitry to send the activation timetable to the host interface circuitry.

2. The system of claim 1 , where the era timescale comprises an annual timescale, a seasonal timescale, a quarterly timescale, a monthly timescale, a weekly timescale, a daily timescale, or any combination thereof.

3. The system of claim 1 , where the activation timetable circuitry is configured to generate, at the prescriptive engine layer, a user interface comprising a graphical representation of at least a portion of the activation timetable.

4. The system of claim 3 , where the user interface allows input of an operator-initiated change to the activation states.

5. The system of claim 4 , where the user interface allows input of the operator-initiated change through operator selection of the activation timeslot.

6. The system of claim 3 , where the user interface comprises an informational panel to display an efficiency differential comparing implementation of the activation threshold to non-implementation of the activation threshold.

7. The system of claim 6 , where the informational panel allows display of a predicted effect on the efficiency differential corresponding to a specific change to the activation threshold.

8. A method comprising:
at network interface circuitry:
receiving historical utilization data for a set of virtual machines;
receiving tagging data defining a functional grouping for a first virtual machine of the set of virtual machines;
receiving consumption metric data including cost metrics for the set of virtual machines; and
receiving a threshold selection value for the functional grouping, the threshold selection value including a target efficiency differential that identifies a difference in ratios of utilization time to non-utilization time for the set of virtual machines and further weights the set of virtual machines based on the consumption metric data; and

at activation timetable circuitry in data communication with the network interface circuitry:
parsing, at an input layer of an activation timetable stack executing on the activation timetable circuitry, the historical utilization data, the consumption metric data, and tagging data to generate time-scaled pattern data;
storing, via a database operation at a data staging layer of the activation timetable stack, the time-scaled pattern data;
accessing, at a transformation layer of the activation timetable stack, the time-scaled pattern data using a memory resource provided by the data staging layer;
identifying, at the transformation layer, an era variable indicative of utilization on an era timescale;
setting, at a configuration layer of the activation timetable stack, an activation threshold for the first virtual machine responsive to the threshold selection value;
passing the era variable from the transformation layer to a data treatment layer of the activation timetable stack;
responsive to the time-scaled pattern data determining an associated value for an era variable for an activation timeslot;
flagging, at a prescriptive engine layer of the activation timetable stack, activation states for the first virtual machine based at least in part on the associated value and the activation threshold;
after flagging the activation states, altering an activation state for the activation timeslot responsive to an adjacent activation state for an adjacent timeslot and an intelligent bucket analysis; and
generating, at the prescriptive engine layer, an activation timetable for the set of virtual machines, the activation timetable defining activation states for the first virtual machine; and

sending, via the network interface circuitry, the activation timetable to a host interface configured to control the activation states for at least the first virtual machine.

9. The method of claim 8 further comprising, generating, at the prescriptive engine layer, a user interface comprising a graphical representation of at least a portion of the activation timetable.

10. The method of claim 9 further comprising receiving, via the user interface, input of an operator-initiated change to the activation states.

11. The method of claim 10 , where receiving the input of the operator-initiated change comprises receiving an operator selection of the activation timeslot.

12. The method of claim 9 , further comprising causing, on an informational panel of the user interface, display of an efficiency differential comparing implementation of the activation threshold to non-implementation of the activation threshold.

13. The method of claim 12 , further comprising causing, on the informational panel, display of a predicted effect on the efficiency differential corresponding to a specific change to the activation threshold.

14. A product comprising:
a non-transitory machine-readable medium; and
instructions stored on the non-transitory machine-readable medium, the instructions configured to, when executed, cause a machine to:
at network interface

Assignees

• Accenture Global Solutions Ltd

Inventors

• Srinivasan Madhan Kumar
• Purushothaman Arun
• Kolachalam Manish Sharma
• Eisenstein Michael S

Classifications

• G06F11/301

  where the computing system is a virtual computing platform, e.g. logically partitioned systems (virtual machines G06F9/45533; logical partitioning of resources G06F9/5077) · CPC title

• H04L67/1029

  using data related to the state of servers by a load balancer · CPC title

• H04L67/1012

  based on compliance of requirements or conditions with available server resources · CPC title

• G06F16/248

  Presentation of query results · CPC title

• G06F11/0712

  in a virtual computing platform, e.g. logically partitioned systems · CPC title