Workload optimization, scheduling, and placement for rack-scale architecture computing systems

Patent metadata

Field	Value
Publication number	US-2016359683-A1
Application number	US-201515114687-A
Country	US
Kind code	A1
Filing date	Feb 24, 2015
Priority date	Feb 27, 2014
Publication date	Dec 8, 2016
Grant date	—

Abstract

Technologies for datacenter management include one or more computing racks each including a rack controller. The rack controller may receive system, performance, or health metrics for the components of the computing rack. The rack controller generates regression models to predict component lifespan and may predict logical machine lifespans based on the lifespan of the included hardware components. The rack controller may generate notifications or schedule maintenance sessions based on remaining component or logical machine lifespans. The rack controller may compose logical machines using components having similar remaining lifespans. In some embodiments the rack controller may validate a service level agreement prior to executing an application based on the probability of component failure. A management interface may generate an interactive visualization of the system state and optimize the datacenter schedule based on optimization rules derived from human input in response to the visualization. Other embodiments are described and claimed.

First claim

1 - 25 . (canceled)

26 . A rack controller of a computing rack, the rack controller comprising:
a performance monitoring module to receive a metric associated with a hardware component of the computing rack, wherein the metric comprises a system metric, a performance metric, or a health metric; and an analytics module to:
determine a regression model for the hardware component based on the metric associated with the hardware component;
determine a mean-time-to-failure value for the hardware component based on the regression model for the hardware component; and
determine a mean-time-to-failure value for a logical machine based on the mean-time-to-failure value for the hardware component, wherein the logical machine is associated with the hardware component.

27 . The rack controller of claim 26 , wherein to receive the metric comprises to receive the metric from a metric component of the hardware component.

28 . The rack controller of claim 26 , wherein the hardware component comprises a compute resource, a memory resource, a storage resource, or a network resource.

29 . The rack controller of claim 26 , wherein to determine the regression model comprises to determine a linear regression model.

30 . The rack controller of claim 26 , wherein to determine the regression model comprises to determine a non-linear regression model.

31 . The rack controller of claim 26 , wherein to determine the mean-time-to-failure value for the hardware component comprises to:
determine a predicted metric associated with the hardware component based on the regression model; and compare the predicted metric to a predefined threshold metric.

32 . The rack controller of claim 26 , wherein to determine the mean-time-to-failure value for the hardware component comprises to:
determine a predicted metric associated with the hardware component based on the regression model; determine a service level metric of a service level agreement associated with the hardware component; and compare the predicted metric to the service level metric.

33 . The rack controller of claim 26 , further comprising a datacenter management module to notify a user of the mean-time-to-failure value for the hardware component.

34 . The rack controller of claim 26 , further comprising a datacenter management module to determine a future time for a maintenance session associated with the logical machine based on the mean-time-to-failure value for the logical machine.

35 . The rack controller of claim 34 , wherein:
the performance monitoring module is further to receive a performance indicator associated with a computing application assigned to the logical machine; and to determine the future time further comprises to determine the future time based on the performance indicator.

36 . The rack controller of claim 26 , further comprising a datacenter management module to:
identify a plurality of hardware components of the computing rack, wherein the plurality of hardware components comprises the hardware component and wherein a mean-time-to-failure value associated with each of the hardware components is similar to the mean-time-to-failure value of the hardware component; and compose the logical machine to include the plurality of hardware components.

37 . A method for datacenter management, the method comprising:
receiving, by a rack controller of a computing rack, a metric associated with a hardware component of the computing rack, wherein the metric comprises a system metric, a performance metric, or a health metric; determining, by the rack controller, a regression model for the hardware component based on the metric associated with the hardware component; determining, by the rack controller, a mean-time-to-failure value for the hardware component based on the regression model for the hardware component; and determining, by the rack controller, a mean-time-to-failure value for a logical machine based on the mean-time-to-failure value for the hardware component, wherein the logical machine is associated with the hardware component.

38 . The method of claim 37 , wherein receiving the metric comprises receiving the metric from a metric component of the hardware component.

39 . The method of claim 37 , wherein determining the mean-time-to-failure value for the hardware component comprises:
determining a predicted metric associated with the hardware component based on the regression model; and comparing the predicted metric to a predefined threshold metric.

40 . The method of claim 37 , wherein determining the mean-time-to-failure value for the hardware component comprises:
determining a predicted metric associated with the hardware component based on the regression model; determining a service level metric of a service level agreement associated with the hardware component; and comparing the predicted metric to the service level metric.

41 . The method of claim 37 , further comprising determining, by the rack controller, a future time for a maintenance session associated with the logical machine based on the mean-time-to-failure value for the logical machine.

42 . The method of claim 41 , further comprising:
receiving, by the rack controller, a performance indicator associated with a computing application assigned to the logical machine; wherein determining the future time further comprises determining the future time based on the performance indicator.

43 . The method of claim 37 , further comprising:
identifying, by the rack controller, a plurality of hardware components of the computing rack, wherein the plurality of hardware components comprises the hardware component and wherein a mean-time-to-failure value associated with each of the hardware components is similar to the mean-time-to-failure value of the hardware component; and composing, by the rack controller, the logical machine to include the plurality of hardware components.

44 . One or more computer-readable storage media comprising a plurality of instructions that in response to being executed cause a rack controller of a computing rack to:
receive a metric associated with a hardware component of the computing rack, wherein the metric comprises a system metric, a performance metric, or a health metric; determine a regression model for the hardware component based on the metric associated with the hardware component; determine a mean-time-to-failure value for the hardware component based on the regression model for the hardware component; and determine a mean-time-to-failure value for a logical machine based on the mean-time-to-failure value for the hardware component, wherein the logical machine is associated with the hardware component.

45 . The one or more computer-readable storage media of claim 44 , wherein to receive the metric comprises to receive the metric from a metric component of the hardware component.

46 . The one or more computer-readable storage media of claim 44 , wherein to determine the mean-time-to-failure value for the hardware component comprises to:
determine a predicted metric associated with the hardware component based on the regression model; and compare the predicted metric to a predefined threshold metric.

47 . The one or more computer-readable storage media of claim 44 , wherein to determine the mean-time-to-failure value for the hardware component comprises to:
determine a predicted metric associated with the hardware component based on the regre

Assignees

• Intel Corp

Inventors

• Bartfai-Walcott Katalin K
• Woods Chris
• Estrada Giovani
• Kennedy John
• Butler Joseph
• Putyrski Slawomir
• Leckey Alexander
• Bayon-Molino Victor M
• Upton Connor
• Metsch Thijs

Classifications

• H04L41/5054

  Automatic deployment of services triggered by the service manager, e.g. service implementation by automatic configuration of network components · CPC title

• H04L41/0883

  Semiautomatic configuration, e.g. proposals from system · CPC title

• H04L43/0817

  by checking functioning · CPC title

• H04L41/22

  comprising specially adapted graphical user interfaces [GUI] · CPC title

• H04L41/5009

  Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF] · CPC title