Priority switching based on resource usage patterns

Patent metadata

Field	Value
Publication number	US-2020028895-A1
Application number	US-201816041299-A
Country	US
Kind code	A1
Filing date	Jul 20, 2018
Priority date	Jul 20, 2018
Publication date	Jan 23, 2020
Grant date	—

Abstract

A processing device in a system can allocate builds to execute jobs across servers in a distributed server environment such as a continuous integration (CI) or continuous deployment (CD) server environment, determine workload patterns, and store a workload pattern vector including a resource usage pattern for the builds and the servers. The processing device can provide an interface that allows a user to switch priorities of user-owned builds in server environments, even after execution of the jobs has begun. The user's ability to prioritize or reprioritize builds is limited to user-prioritizable builds as determined based on the resource usage patterns of the builds running the jobs. The user can thus prioritize some user-owned builds without adversely affecting builds of other users.

First claim

1 . A system comprising:
a processing device; and a memory device including instructions that are executable by the processing device for causing the processing device to perform operations comprising:
allocating a plurality of jobs across a plurality of servers, the plurality of jobs involving execution of a plurality of builds on the plurality of servers;
storing a workload pattern vector that includes a plurality of resource usage patterns corresponding to the plurality of jobs and the plurality of builds, each resource usage pattern indicating how computing resources are consumed by a server to execute a respective build from among the plurality of builds;
determining, using the workload pattern vector, a plurality of statistical indicators corresponding to the plurality of jobs, each respective statistical indicator representing a statistical measure of resource usage for the plurality of builds for a respective job among the plurality of jobs; and
displaying, based at least in part on the plurality of statistical indicators, a list of builds including one or more user-prioritizable builds from among the plurality of builds.

2 . The system of claim 1 wherein the statistical indicator comprises a standard deviation of a median of statistical distance between the resource usage patterns of at least one set of builds for each job.

3 . The system of claim 2 wherein the at least one set of builds comprises at least one of, all builds, all builds except an earliest build, or all builds except two earliest builds.

4 . The system of claim 2 wherein the list of builds is displayed at least in part by comparing the standard deviation of the median of statistical distance to a threshold value.

5 . The system of claim 2 wherein the median of statistical distance is determined based on sparse data from the workload pattern vector.

6 . The system of claim 1 wherein the operations further comprise:
receiving user input directed to prioritizing at least one of the user-prioritizable builds; and
prioritizing the at least one of the user-prioritizable builds in response to the user input.

7 . The system of claim 1 wherein the plurality of builds is allocated across the plurality of servers based at least in part on a hardware specification for the plurality of servers and a number of simultaneous builds permitted on each of the plurality of servers.

8 . A method comprising:
allocating, by a processing device, a plurality of jobs across a plurality of servers, the plurality of jobs involving execution of a plurality of builds on the plurality of servers; storing, in a storage device, a workload pattern vector that includes a plurality of resource usage patterns corresponding to the plurality of jobs and the plurality of builds, each resource usage pattern indicating how computing resources are consumed by a server to execute a respective build from among the plurality of builds; determining, by the processing device, using the workload pattern vector, a plurality of statistical indicators corresponding to the plurality of jobs, each respective statistical indicator representing a statistical measure of resource usage for the plurality of builds for a respective job among the plurality of jobs; and displaying, by the processing device, based at least in part on the plurality of statistical indicators, a list of builds including one or more user-prioritizable builds from among the plurality of builds.

9 . The method of claim 8 wherein the statistical indicator comprises a standard deviation of a median of statistical distance between the resource usage of at least one set of builds for each job.

10 . The method of claim 9 wherein the at least one set of builds comprises at least one of, all builds, all builds except an earliest build, or all builds except two earliest builds.

11 . The method of claim 9 wherein the list of builds is displayed at least in part by comparing the standard deviation of the median of statistical distance to a threshold value.

12 . The method of claim 9 wherein the median of statistical distance is determined based on sparse data from the workload pattern vector.

13 . The method of claim 8 further comprising:
receiving user input directed to prioritizing at least one of the user-prioritizable builds; and
prioritizing the at least one of the user-prioritizable builds in response to the user input.

14 . The method of claim 8 wherein the plurality of builds is allocated across the plurality of servers based at least in part on a hardware specification for the plurality of servers and a number of simultaneous builds permitted on each of the plurality of servers.

15 . A non-transitory computer-readable medium comprising program code that is executable by a processing device for causing the processing device to:
allocate a plurality of jobs across a plurality of servers, the plurality of jobs involving execution of a plurality of builds on the plurality of servers; store a workload pattern vector that includes a plurality of resource usage patterns corresponding to the plurality of jobs and the plurality of builds, each resource usage pattern indicating how computing resources are consumed by a server to execute a respective build from among the plurality of builds; determine, using the workload pattern vector, a plurality of statistical indicators corresponding to the plurality of jobs, each respective statistical indicator representing a statistical measure of resource usage for the plurality of builds for a respective job among the plurality of jobs; and display, based at least in part on the plurality of statistical indicators, a list of builds including one or more user-prioritizable builds from among the plurality of builds.

16 . The non-transitory computer-readable medium of claim 15 wherein the statistical indicator comprises a standard deviation of a median of statistical distance between the resource usage of at least one set of builds for each job.

17 . The non-transitory computer-readable medium of claim 16 wherein the list of builds is displayed at least in part by comparing the standard deviation of the median of statistical distance to a threshold value.

18 . The non-transitory computer-readable medium of claim 16 wherein the median of the statistical distance is determined based on sparse data from the workload pattern vector.

19 . The non-transitory computer-readable medium of claim 15 further comprising program code that is executable by the processing device for causing the processing device to:
receive user input directed to prioritizing at least one of the user-prioritizable builds; and
prioritize the at least one of the user-prioritizable builds in response to the user input.

20 . The non-transitory computer-readable medium of claim 15 wherein the plurality of builds is allocated across the plurality of servers based at least in part on a hardware specification for the plurality of servers and a number of simultaneous builds permitted on each of the plurality of servers.

Assignees

• Red Hat Israel Ltd

Inventors

• Bregman Arie
• Sariel David

Classifications

• H04L67/1004 Primary

  Server selection for load balancing · CPC title

• G06F9/505

  considering the load · CPC title

• G06F9/5083

  Techniques for rebalancing the load in a distributed system · CPC title