Granted Patent B2

US 11,513,862 · App. 16/877,253 · Granted Nov 29, 2022

Inventors: Tandava Venkata Krishna Popuri (Bengaluru, IN); Vaideeswaran Ganesan (Bengaluru, IN); Pravin Janakiram (Bangalore, IN); Balaji Bondhili Singh (Bangalore, IN); Suren Kumar (Vellore, IN)

Assignee: Dell Products L.P.

G06F9/505G06F9/485G06F9/5005G06F9/5038G06F2209/508

Abstract

Claims (67)

1. A deployment manager, comprising:

storage for storing:

a state repository comprising a plurality of state transitions associated with event descriptions generated by a computing device, wherein the event descriptions comprise alerts generated by components of the computing device, wherein the alerts comprise noun-verb phrases that describe characteristics of at least one of the components; and

a computing device manager programmed to:

obtain an event description of the event descriptions;

identify a component of the components associated with the event description;

divide the event description into one or more portions that each include only noun or verb phrases;

analyze at least one of the portions to identify a computing device specific meaning of the at least one of the portions;

generate a state transition of the state transitions using an identifier of the component and the computing device specific meaning;

obtain a new event description associated with:

the computing device, and

a workload performed by the computing device;

in response to obtaining the new event description:

match the new event description to the state transition; and

manage the workload based on a predicted next state associated with the state transition.

2. The deployment manager of claim 1 , wherein generating the state transition of the state transitions using the identifier of the component and the computing device specific meaning comprises:

identifying a next state associated with the computing device specific meaning; and

associating the identifier of the component, the computing device specific meaning, and the next state.

3. The deployment manager of claim 2 , wherein the computing device manager is further programmed to:

a first likelihood that the computing device will transition to the next state when the component is associated with an occurrence of an event that is associated with the computing device specific meaning, and

a second likelihood that the computing device will transition to a second next state when the component is associated with the occurrence of the event that is associated with the computing device specific meaning.

4. The deployment manager of claim 1 , wherein managing the workload based on a predicted next state associated with the state transition comprises:

obtain a second new event description associated with:

the computing device, and

the workload performed by the computing device;

in response to obtaining the second new event description:

make a determination that the second new event description does not match any state transition of the plurality of state transitions; and

discard the second new event description without managing the workload.

5. The deployment manager of claim 1 , wherein the computing device manager is programmed to discard at least 99% of event descriptions generated by the computing device without determining that a state of the computing device is likely to change.

6. The deployment manager of claim 1 , wherein the computing device manager is programmed to add a new state transition to the state transitions when any event description that is obtained does not match any of the state transitions.

7. The deployment manager of claim 1 , wherein the state transition specifies that the computing device is more likely to transition to the next state than a second next state.

8. The deployment manager of claim 1 , wherein the state transition further specifies a transition time associated with the next state and a second transition time associated with the second next state.

9. The deployment manager of claim 1 , wherein the new event description comprises a human readable description of a computing device event.

10. The deployment manager of claim 1 , where managing the workload based on the predicted next state associated with the state transition comprises:

terminating the workload;

instantiating a new instance of the workload hosted by the computing device; and

modifying an operation of the workload to avoid the predicted next state.

11. The deployment manager of claim 1 , where managing the workload based on the predicted next state associated with the state transition comprises:

allocating additional computing resources to the workload, wherein an amount of the additional computing resources is based on a transition time associated with the predicted next state.

12. A method for managing a computing device using a plurality of state transitions associated with event descriptions generated by the computing device, comprising:

obtaining an event description of the event descriptions, wherein the event descriptions comprise alerts generated by components of the computing device, wherein the alerts comprise noun-verb phrases that describe characteristics of at least one of the components;

identifying a component of the components associated with the event description;

dividing the event description into one or more portions that each include only noun or verb phrases;

analyzing at least one of the portions to identify a computing device specific meaning of the at least one of the portions;

generating a state transition of the state transitions using an identifier of the component and the computing device specific meaning;

obtaining a new event description associated with:

the computing device, and

a workload performed by the computing device;

in response to obtaining the new event description:

matching the new event description to the state transition; and

managing the workload based on a predicted next state associated with the state transition.

13. The method of claim 12 , wherein the state transition specifies that the computing device is more likely to transition to the next state than a second next state.

14. The method of claim 12 , wherein the state transition further specifies a transition time associated with the next state and a second transition time associated with the second next state.

15. A non-transitory computer readable medium comprising computer readable program code, which when executed by a computer processor enables the computer processor to perform a method for managing a computing device using a plurality of state transitions associated with event descriptions generated by the computing device, the method comprising:

obtaining an event description of the event descriptions, wherein the event descriptions comprise alerts generated by components of the computing device, wherein the alerts comprise noun-verb phrases that describe characteristics of at least one of the components;

identifying a component of the components associated with the event description;

dividing the event description into one or more portions that each include only noun or verb phrases;

analyzing at least one of the portions to identify a computing device specific meaning of the at least one of the portions;

generating a state transition of the state transitions using an identifier of the component and the computing device specific meaning;

obtaining a new event description associated with:

the computing device, and

a workload performed by the computing device;

in response to obtaining the new event description:

matching the new event description to the state transition; and

managing the workload based on a predicted next state associated with the state transition.

16. The non-transitory computer readable medium of claim 15 , wherein the state transition specifies that the computing device is more likely to transition to the next state than a second next state.

17. The non-transitory computer readable medium of claim 15 , wherein the state transition further specifies a transition time associated with the next state and a second transition time associated with the second next state.

Assignments (9)