Probabilistic classifying system and method for a distributed computing environment
View Patent ↗A probabilistic classifying system for a distributed computing environment includes a computer-executable system for obtaining multiple measured values associated with the operation of a plurality of resources of a distributed computing environment. The system uses the measured values to generate a health metric assessment using a probabilistic classifier in which the health metric assessment comprising one of multiple finite health metric assessments. The probabilistic classifier generates the health metric assessment according to the obtained measured values and a statistical inference upon at least a portion of a plurality of previously obtained measured values of the resources of the distributed computing environment.
1. A probabilistic classifying system for a distributed computing environment, the system comprising:
a computing system comprising at least one processor and at least one memory to store instructions that are executed by the at least one processor to:
train a probabilistic model using machine learning by using one of predefined health values from a manufacturer of the distributed computing environment and a customer survey obtained from a plurality of users of the distributed computing environment to generate the probabilistic model;
obtain in real-time a plurality of measured values associated with operation of a plurality of resources of a distributed computing environment;
convert and normalize each of the plurality of measured values having a first, disparate format to a second, common format that may analyzed by the probabilistic classifying system;
generate a health metric assessment using a probabilistic classifier, the health metric assessment comprising one of a plurality of finite health metric assessments, wherein the probabilistic classifier generates the health metric assessment according to the obtained plurality of measured values, the probabilistic model, and a statistical inference upon at least a portion of a plurality of previously obtained measured values of the resources of the distributed computing environment having the second, common format; and
perform at least one remedial action when the health metric assessment exceeds a specified threshold, the at least one remedial action comprising displaying a message on a user interface, displaying a selectable button on the user interface that alerts that the health metric assessment exceeds the specified threshold, and automatically modifying at least one particular resource of the plurality of resources of the distributed computing environment.
2. The probabilistic classifying system of claim 1 , wherein the instructions are further executed to perform a training operation for the probabilistic classifying tool using the previously obtained measured values to generate the statistical inference.
3. The probabilistic classifying system of claim 2 , wherein the previously obtained measured values comprise simulated values, wherein the training operation is conducted following manufacture of the distributed computing environment and before any applications are deployed on the distributed computing environment.
4. The probabilistic classifying system of claim 1 , wherein the measured values comprise at least one of a processing load of one of the resources, a communication throughput of one of the resources, a memory usage of one of the resources, an ambient temperature that the resource is currently operating in, and a humidity that the resource is currently operating in.
5. The probabilistic classifying system of claim 1 , wherein the probabilistic classifier comprises a naïve Bayes probabilistic classifier.
6. The probabilistic classifying system of claim 1 , wherein the instructions are further executed to obtain a health metric assessment for a portion of the resources of the distributed computing environment, wherein the portion of the resources are associated with an application executed on the distributed computing environment.
7. The probabilistic classifying system of claim 1 , wherein the instructions are further executed to generate the health metric assessment on an ongoing basis.
8. A probabilistic classifying method comprising:
training, using at least one processor, a probabilistic model using machine learning by using one of predefined health values from a manufacturer of a distributed computing environment and a customer survey obtained from a plurality of users of the distributed computing environment to generate the probabilistic model;
obtaining in real-time, using an operations management application stored on a non-transitory medium and executed by the at least one processor, a plurality of measured values associated with operation of a plurality of resources of a distributed computing environment;
converting and normalizing, using the at least one processor, each of the plurality of measured values having a first, disparate format to a second, common format that may analyzed by the at least one processor;
generating, using the at least one processor, a health metric assessment using a probabilistic classifier, the health metric assessment comprising one of a plurality of finite health metric assessments, wherein the probabilistic classifier generating the health metric assessment according to the obtained plurality of measured values, the probabilistic model, and a statistical inference upon at least a portion of a plurality of previously obtained measured values of the resources of the distributed computing environment having the second, common format; and
performing, using the at least one processor, at least one remedial action when the health metric assessment exceeds a specified threshold, the at least one remedial action comprising displaying a message on a user interface, displaying a selectable button on the user interface that alerts that the health metric assessment exceeds the specified threshold, and automatically modifying at least one particular resource of the plurality of resources of the distributed computing environment.
9. The probabilistic classifying method of claim 8 , further comprising performing a training operation using the previously obtained measured values to generate the statistical inference.
10. The probabilistic classifying method of claim 9 , further comprising conducting the training operation following manufacture of the distributed computing environment and before any applications are deployed on the distributed computing environment, wherein the previously obtained measured values comprise simulated values.
11. The probabilistic classifying method of claim 8 , wherein the probabilistic classifier comprises a naïve Bayes probabilistic classifier.
12. The probabilistic classifying method of claim 8 , further comprising obtaining the health metric assessment for a portion of the resources of the distributed computing environment, wherein the portion of the resources are associated with an application executed on the distributed computing environment.
13. The probabilistic classifying method of claim 8 , further comprising generating the health metric assessment on an ongoing basis.
14. A non-transitory, computer readable storage medium having instructions stored thereon that when executed by at least one processor, cause the at least one processor to perform operations comprising:
training a probabilistic model using machine learning by using one of predefined health values from a manufacturer of a distributed computing environment and a customer survey obtained from a plurality of users of the distributed computing environment to generate the probabilistic model;
obtaining in real-time a plurality of measured values associated with operation of a plurality of resources of a distributed computing environment;
converting and normalizing each of the plurality of measured values having a first, disparate format to a second, common format that may analyzed by the at least one processor;
generating a health metric assessment using a probabilistic classifier, the health metric assessment comprising one of a plurality of finite health metric assessments, wherein the probabilistic classifier generating the health metric assessment according to the obtained plurality of measured values, the probabilistic model, and a statistical inference upon at least a portion of a plurality of previously obtained measured values of the resources of the distributed computing environment having the second, common format; and
performing at least one remedial action when the health metric assessment exceeds a specified threshold, the at least one remedial action comprising displaying a message on a user interface, displaying a selectable button on the user interface that alerts that the health metric assessment exceeds the specified threshold, and automatically modifying at least one particular resource of the plurality of resources of the distributed computing environment.
15. The non-transitory, computer readable storage medium of claim 14 , wherein the probabilistic classifier comprises a naïve Bayes probabilistic classifier.
16. The non-transitory, computer readable storage medium of claim 14 , the operations further comprising obtaining the health metric assessment for a portion of the resources of the distributed computing environment, wherein the portion of the resources are associated with an application executed on the distributed computing environment.
17. The non-transitory, computer readable storage medium of claim 14 , the operations further comprising generating the health metric assessment on an ongoing basis.