Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

A method and system of performance governance for Cloud Computing based on Event-driven mechanism

IP.com Disclosure Number: IPCOM000240110D
Publication Date: 2015-Jan-04
Document File: 6 page(s) / 140K

Publishing Venue

The IP.com Prior Art Database

Abstract

With the prevalence of Cloud service and Cloud computing, more server nodes are added to the system, during the development or production environment, it's very normal and frequent that a wide scope of system changes will be applied to the Cloud infrastructure, changes such as : hotfix, the configuration change including the CPU/RAM or any specific application (middle-ware, database parameters) tuning, etc. So It's becoming more and more complex to track and evaluate the performance impact for single or even multiple system change, and it's difficult to analyse the performance change regarding this by traditional approach. In order to detect the performance degradation/improvement as early as possible, a proactive-detection mechanism will be implemented, whenever and whatever event the system occurs(configuration change, code change, etc.), the exact events info will be collected, tracked, analysed, simply, a core idea is: A predefined performance change set is defined for each event definition and will be refined by the Learning System. The Event-Performance Analyzer will identify the correlation between the event and performance based on Correlation, Pattern.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 28% of the total text.

Page 01 of 6

A method and system of performance governance for Cloud Computing based on Event

A method and system of performance governance for Cloud Computing based on Event-

With the prevalence of Cloud service and Cloud computing, more and more services are emerging on the Cloud managed service systems, more server nodes are added to the system, in the development or production environment, it's very normal and frequent that a wide scope of system changes will be applied to the Cloud infrastructure for the purpose of hot fix, upgrade or performance tuning, e.g. the code changes (hot fix) to the various service components, the configuration change including the CPU/RAM or any specific application (middle-ware, database parameters) tuning, etc.

It's becoming more and more complex to track and evaluate the performance impact for single or even multiple system changes since these changes can come from the various sources(front-end, middle-ware, hardware ,or back-end system), and because of the resource sharing and workload dynamic changes in Cloud computing environment, so it's difficult to analyze the performance change regarding this by traditional approach.

Cloud computing has been brought to real world in recent years. There are many types of implementations for IaaS, PaaS, SaaS or BPaaS etc. It's very important to govern performance of cloud computing systems whatever types they are. All cloud computing vendors provide certain methods/tools to do this. But it's still very hard to get it done accurately, comprehensively and efficiently.

There are plenty of events occurring in cloud computing systems. There are some events included but not limited to: 1, Events from central/managing side:

Modifications to software/middle-ware/hardware, such as adding CPU, Memory, Hard disk, modifying parameters of database configuration, installing a fix pack, upgrading software or plug-in etc..

2, Operations from managing side:

Customer 1 start/stop/modify VMs(virtual machines), a heap dump event occurs in Customer 2's VM which used as an application server, a user from Customer 3 start to execute a heavy workload from VM3...

3, Unexpected events:

Disaster in storage system, unexpected disk full, out of memory etc.

All these 'events' will contribute to overall system's performance and all end-users longing the system will be affected. To keep system's healthy status and normal use is a key job for cloud computing system provider/creator and also to the users.

A novel idea is to track and measure all the changes by a predefined matrix while monitoring all applications and systems ' status based on the time-line, and then recognize the "system change" vs. "performance change" pattern. By learning the difference between initial judgment and actual correlation-ship to refine the change matrix and the actual changes vs. performance pattern so that this system can make more accurate change vs. performance measurement, and performance learning and prediction.

1

-...