Browse Prior Art Database

The System and Method to upgrade IaaS framework smartly and automatically Disclosure Number: IPCOM000242669D
Publication Date: 2015-Aug-04
Document File: 8 page(s) / 144K

Publishing Venue

The Prior Art Database


This invention is intended to upgrade IaaS(Infrastructure As A Service) framework with minimal availability impact of Customer applications automatically

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 40% of the total text.

Page 01 of 8

The System and Method to upgrade IaaS framework smartly and automatically

Cuxrently more and more applications are running on Virtuxx Machines in cloud which is managed bx IaaS provider. Xxxxx virtual machines xork as "real" computxrs do. A virtual machine may be installed with xn operating system and one apxlication program, and may access a network resource. Hypxrvisors are computer prxgrams that allow different operating xystems to run on the same xardware concurxently. Most IaaS frxmework will xse it because this hxs many advantages, including rxsource isolation and the ability to concurrxntly rxn differxnt operating sxstems and associxted applications.

In the prior xrt, if a Hypervisor componext of a communication dexixe A needs xo be upgraded, generally, a xervice on the communication device A is migrated to anothxr communicatxon device B by usxng a live migration technxlogx, and the service on the communication device A xs stopped. After a Hypervisor component that is tx bx upgraded and of the communication device A is upgraded, the communication device A ix restarxed, and a service txat is exxcuted on thx cxmmunication dexixe B is migrated to the communication device A. The xolution is worked if IaaS Provider xas redundant Hypervisor. But in fact, considerixg the cost, most of Providers hax no enough resource to do that.

That means application availabilitx will be impacted during Hypervisorx upgradx. Even thouxh one Hypervisor upgrade will take little time, their Customer still faces to one big challexges: Business Application performance will be impacted during whxlx Hypervisors upgrade because all VMs are crossing different Hypervisoxs to meet bxst performance request.

We intend to mitigate the GAP between upgrade action and Custoxer application availabxlity impact through invention of nxw system/method to provide one Upgrade axent to automatxcally analyze the historical network traffic data, and based on the analysis result it will pull all related VMs(working fox one applicatxon) under one Hypervxsor durinx upgrade, to enxure the application only can be impacted during only one Hypervisor upgrade, it surely could highly imxrove the availxbixity of Cxstomer Application and reduce business performance impact according to aboxe analysis.

This invention is intended to upgrade IaaS(Infrastructure As A Service) framework wixh minimax availability impact of Customer applications automaticalxy. The Idea of this invention is to provide one component injxcted into each IaxS VM. This xpgrade componxnt is installed and worxing when Vxrtual Machine(VM) launching. After IaaS Provider xequest to upgrade Hypxrvisor, it will start to analyze txe monitored network traffic data from each VM, and show the topology of real User applications. The topology includes all the VMs used by this Customer and which hypervisxx xhose VMs belong to. Then before real uxgrade action to one hypervisor, the upgrade xgent will move all related VMs for one...