Cloud service tuning

Patent metadata

Field	Value
Publication number	US-2017331907-A1
Application number	US-201515528277-A
Country	US
Kind code	A1
Filing date	Jun 30, 2015
Priority date	Dec 12, 2014
Publication date	Nov 16, 2017
Grant date	—

Abstract

According to an example, to tune a cloud service, an event and an event type on a cloud server application are detected by monitoring at least one of a hardware value and a software value. An application parameter to tune is determined based on the event type, and a tuning priority of the application parameter based on historical tuning data is fetched. The application parameter is tuned by modifying a software variable. In the event that tuning the application parameter resolves the event type, the tuning priority of the parameter is increased. In the event that tuning the parameter fails to resolve the event type, the parameter priority is decreased, the tuning is rolled back, and a default tune setting is invoked.

First claim

What is claimed is:

1 . A method of tuning a cloud service, comprising:
detecting an event and an event type on a cloud server application by monitoring at least one of a hardware value and a software value on a cloud server; determining at least one exposed application parameter to tune based on the event type; determining a tuning priority of the at least one application parameter based on historical tuning data; automatically tuning the at least one application parameter by modifying a software variable on the cloud server; and in the event that tuning the at least one application parameter resolves the event type, increasing the tuning priority of the at least one application parameter, and in the event that tuning the at least one application parameter fails to resolve the event type, decreasing the priority of the at least one application parameter, rolling back the tuning of the at least one application parameter, and transmitting an instruction to invoke a default tune setting.

2 . The method according to claim 1 , wherein automatically tuning the at least one application parameter further comprises looping through an order of priority.

3 . The method according to claim 1 , wherein the event type comprises a performance degradation alert.

4 . The method according to claim 1 , wherein the event type comprises a threshold level alert.

5 . The method according to claim 1 , wherein the at least one application parameter comprises at least one of a database connection size, a caching value, a connection pool thread value, and a worker threads value.

6 . The method according to claim 1 , wherein automatically tuning the application parameter comprises tuning the application parameter incrementally.

7 . The method according to claim 1 , where modifying a software variable on the cloud server comprises at least one of increasing a variable, decreasing a variable, and toggling a variable.

8 . The method according to claim 1 , wherein the default tune setting comprises an auto scaling function.

9 . A system for tuning a cloud service, the system comprising a processing resource in communication with a network interface and a computer readable medium, wherein the computer readable medium contains a set of instructions and wherein a processing unit is designed to carry out a set of instructions to:
detect, on the processing resource, a first monitoring event and an event type on a cloud server application; determine at least one application parameter exposed in the cloud server application; determine an order of priority for tuning the at least one application parameter based on the detected event type; tune the application parameter; detect, on the processing resource, a second monitoring event on the cloud server application; and determine whether the second monitoring event indicates that the first monitoring event has been resolved, wherein in the event that the first monitoring event is not resolved, the processing resource is to decrease the priority of the at least one application parameter, roll back the tuned application parameter to a previous state, and invoke an auto-scaling function.

10 . The system according to claim 9 , wherein the first monitoring event comprises at least one of a CPU alert, a memory alert, a disk drive alert, a bandwidth alert, and an application alert.

11 . The system according to claim 9 , wherein the previous state comprises a virtual machine image.

12 . The system according to claim 9 , wherein the instructions are to detect, on the processing resource, the second monitoring event on the cloud server application at a pre-set time after the application parameter is tuned.

13 . A non-transitory computer readable storage medium on which is embedded a computer program, said computer program to deploy a tunable cloud services application, said computer program comprising a set of instructions to:
associate at least one tunable application parameter with at least one event type; expose the at least one tunable application parameter to a production cloud server for auto-remediation on the production cloud server; set an adjustable priority level for the at least one application parameter; set a rollback point for the at least one application parameter; set a monitor threshold for the at least one event type; set a default tune action; and deploy the tunable cloud services application to the production cloud server.

14 . The computer readable storage medium according to claim 13 , wherein the event type is based on an application response time.

15 . The computer readable storage medium according to claim 13 , wherein the default tune action comprises an instruction to increase the resources available to the production cloud server.

Assignees

• Hewlett Packard Entpr Dev Lp

Inventors

• Jagannath Kishore
• Srinivas Santhosh

Classifications

• H04L43/0817

  by checking functioning · CPC title

• H04L41/0823

  characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability (for optimising operational conditions of wireless networks H04W24/02) · CPC title

• H04L43/16

  Threshold monitoring · CPC title

• H04L67/10

  in which an application is distributed across nodes in the network (software deployment G06F8/60; multiprogramming arrangements G06F9/46) · CPC title

• H04L67/16 Primary

  Electricity · mapped topic