Real-time analytics based monitoring and classification of jobs for a data processing platform

• US 10,216,543 B2
• Filed: 08/11/2016
• Issued: 02/26/2019
• Est. Priority Date: 08/14/2015
• Status: Active Grant

• Current Assignee: Mitylytics Inc.
• Original Assignee: Mitylytics Inc.
• Inventors: Singh, Maneesh Kumar, Divate, Hrishikesh Pradeep
• Primary Examiner(s): Kamran, Mehran
• Application Number:
  US15/234,637
• Publication Number:
  US 20170046203A1
• Time in Patent Office: 929 Days
• Field of Search: 718104
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system
• CPC Class Codes: G06N 10/00 Quantum computing, i.e. information processing based on quantum-mechanical phenomena
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

In one embodiment, a method selects a new job to schedule for execution on a data processing system. The new job includes a. Performance information for a set of current jobs that are being executed in the data processing system is retrieved where the set of jobs are assigned to queues and currently classified with a current classification. The method analyzes the performance information to determine when one or more current jobs in the set of current jobs should be re-classified due to resource usage of a respective current job when being executed in the data processing system and re-classifies the classifications for the one or more current jobs in the queues. Then, the new job is assigned to one of the queues based on the classification of the new job and the classifications of jobs in the queues including the re-classified classifications for the one or more current jobs.

Claims

1. A method comprising:

selecting, by a computing device, a new job to schedule for execution on a data processing system, the new job including a classification in a plurality of classifications, wherein the classification is determined by;

using a process to analyze a set of operations to determine which operation is to be used to classify the current job, wherein a first operation in the set of operations is selected; and

classifying the first operation in a first classification based on resource usage for the first operation, wherein the first classification is determined based on a resource being used by the first operation in a highest percentage usage in the data processing platform compared to other resources used by the first operation;

retrieving, by the computing device, performance information for a set of current jobs that are being executed in the data processing system, wherein the set of jobs are assigned to a plurality of queues and currently classified with a current classification in the plurality of classifications;

analyzing, by the computing device, the performance information to determine when one or more current jobs in the set of current jobs should be re-classified due to resource usage of a respective current job when being executed in the data processing system, wherein analyzing comprises;

determining a second operation in the set of operations; and

determining that the first classification should be changed to the second classification when a resource being used by the second operation has a higher percentage usage in the data processing platform compared to the highest percentage usage for the first operation;

re-classifying, by the computing device, the classifications for the one or more current jobs in the plurality of queues, wherein the first classification is re-classified to the second classification for the first operation; and

assigning, by the computing device, the new job to one of the queues based on the classification of the new job and the classifications of jobs in the plurality of queues including the re-classified classifications for the one or more current jobs.

2. The method of claim 1, further comprising:

receiving the performance information from a plurality of agents installed on data processing devices in the data processing system, the plurality of agents associated resource usage on the data processing devices with current jobs being executed on respective data processing devices.

3. The method of claim 2, wherein the performance information is analyzed to generate resource usage for the plurality of classifications per current job.

4. The method of claim 1, further comprising:

selecting a classification for the new job based on job characteristics associated with the new job.

5. The method of claim 4, wherein selecting the classification comprises:

training a predictor to predict classifications; and

inputting the job characteristics into the predictor to automatically predict the classification for the new job.

6. The method of claim 1, wherein analyzing the performance information comprises:

training a predictor to predict classifications for the set of current jobs; and

inputting the performance information into the predictor to automatically predict the classification for the set of current jobs.

7. The method of claim 1, wherein analyzing the performance information comprises:

applying a first weight to the resource usage for the first operation; and

applying a second weight to the resource usage for the second operation, wherein the first weight is higher than the second weight.

8. The method of claim 1, wherein assigning the new job comprises:

determining which classifications are included in each of the plurality of queues for current jobs in each of the plurality of queues; and

assigning the new job to the one of the queues based on an open slot in the one of the queues having a same classification as the new job.

9. The method of claim 1, further comprising:

assigning classifications to open slots in the plurality of queues based on the re-classifying, the assigning of the classifications to the open slots based on a balancing algorithm of the plurality of classifications in the plurality of queues.

10. The method of claim 1, further comprising:

changing a current job in a first queue in the plurality of queues to a second queue in the plurality of queues after the re-classifying, wherein the changing is performed when the second queue includes less jobs with the classification for the current queue than the first queue.

11. A non-transitory computer-readable storage medium containing instructions, that when executed, control a computer system to be configured for:

selecting a new job to schedule for execution on a data processing system, the new job including a classification in a plurality of classifications, wherein the classification is determined by;

using a process to analyze a set of operations to determine which operation is to be used to classify the current job, wherein a first operation in the set of operations is selected; and

classifying the first operation in a first classification based on resource usage for the first operation, wherein the first classification is determined based on a resource being used by the first operation in a highest percentage usage in the data processing platform compared to other resources used by the first operation;

retrieving performance information for a set of current jobs that are being executed in the data processing system, wherein the set of jobs are assigned to a plurality of queues and currently classified with a current classification in the plurality of classifications;

analyzing the performance information to determine when one or more current jobs in the set of current jobs should be re-classified due to resource usage of a respective current job when being executed in the data processing system, wherein analyzing comprises;

determining a second operation in the set of operations; and

determining that the first classification should be changed to the second classification when a resource being used by the second operation has a higher percentage usage in the data processing platform compared to the highest percentage usage for the first operation;

re-classifying the classifications for the one or more current jobs in the plurality of queues, wherein the first classification is re-classified to the second classification for the first operation; and

assigning the new job to one of the queues based on the classification of the new job and the classifications of jobs in the plurality of queues including the re-classified classifications for the one or more current jobs.

12. The non-transitory computer-readable storage medium of claim 11, wherein analyzing the performance information comprises:

applying a first weight to the resource usage for the first operation; and

applying a second weight to the resource usage for the second operation, wherein the first weight is higher than the second weight.

13. The non-transitory computer-readable storage medium of claim 11, wherein assigning the new job comprises:

determining which classifications are included in each of the plurality of queues for current jobs in each of the plurality of queues; and

assigning the new job to the one of the queues based on an open slot in the one of the queues having a same classification as the new job.

14. The non-transitory computer-readable storage medium of claim 11, wherein analyzing the performance information comprises:

determining a set of operations being performed by a current job;

using a process to analyze the set of operations to determine which operation is to be used to classify the current job; and

classifying the current job based on resource usage for the determined operation.

15. The non-transitory computer-readable storage medium of claim 11, further configured for:

assigning classifications to open slots in the plurality of queues based on the re-classifying, the assigning of the classifications to the open slots based on a balancing algorithm of the plurality of classifications in the plurality of queues.

16. The non-transitory computer-readable storage medium of claim 11, further configured for:

changing a current job in a first queue in the plurality of queues to a second queue in the plurality of queues after the re-classifying, wherein the changing is performed when the second queue includes less jobs with the classification for the current queue than the first queue.

17. The non-transitory computer-readable storage medium of claim 11, further configured for:

selecting a classification for the new job based on job characteristics associated with the new job.

18. The non-transitory computer-readable storage medium of claim 11, wherein analyzing the performance information comprises:

training a predictor to predict classifications for the set of current jobs; and

inputting the performance information into the predictor to automatically predict the classification for the set of current jobs.

19. The non-transitory computer-readable storage medium of claim 11, further configured for:

receiving the performance information from a plurality of agents installed on data processing devices in the data processing system, the plurality of agents associated resource usage on the data processing devices with current jobs being executed on respective data processing devices.

20. An apparatus comprising:

one or more computer processors; and

a non-transitory computer-readable storage medium comprising instructions, that when executed, control the one or more computer processors to be configured for;

selecting a new job to schedule for execution on a data processing system, the new job including a classification in a plurality of classifications, wherein the classification is determined by;

using a process to analyze a set of operations to determine which operation is to be used to classify the current job, wherein a first operation in the set of operations is selected; and

classifying the first operation in a first classification based on resource usage for the first operation, wherein the first classification is determined based on a resource being used by the first operation in a highest percentage usage in the data processing platform compared to other resources used by the first operation;

retrieving performance information for a set of current jobs that are being executed in the data processing system, wherein the set of jobs are assigned to a plurality of queues and currently classified with a current classification in the plurality of classifications;

analyzing the performance information to determine when one or more current jobs in the set of current jobs should be re-classified due to resource usage of a respective current job when being executed in the data processing system, wherein analyzing comprises;

determining a second operation in the set of operations; and

determining that the first classification should be changed to the second classification when a resource being used by the second operation has a higher percentage usage in the data processing platform compared to the highest percentage usage for the first operation;

re-classifying the classifications for the one or more current jobs in the plurality of queues, wherein the first classification is re-classified to the second classification for the first operation; and

assigning the new job to one of the queues based on the classification of the new job and the classifications of jobs in the plurality of queues including the re-classified classifications for the one or more current jobs.