Profoss / Events / January 2008: Virtualisation / Speakers / Matthias Rechenburg

Matthias Rechenburg

Speaker Details
Name Matthias Rechenburg
Title OpenQRM Project Manager
Company
Talks

Matthias Rechenburg is project manager of the openQRM project. Since many years he is involved in all kinds of data-center related open-source projects like high-performance and high-availability clustering, consolidation, network and enterprise storage management. Currently, his most serious interests are about virtualization technologies, their features and capabilities and integration by a unified virtualization layer. He lives in Bonn, Germany, and is working as a freelancer developing for Qlusters. Mostly, he enjoys to code in his home-lab but also likes traveling, meeting other Linux-people and joining all kinds of Linux-related events and congresses.

Interview

Who is OpenQRM for, and what problems does it tackle?

openQRM is an open-source project located at sourceforge which provides a pluggable data-center management platform to automated resource-planning, deployment, monitoring, high-availability and virtualization administration. It's goal is to conform all typical system-administrator tasks within a single management console. openQRM is designed using a pluggable architecture so that third-party tools can be integrated easily by additional plugins.

How was it started and by who?

openQRM is based on a proven commercial project ("QRM" -> Qlusters Resource Manager) which was open-sourced by Qlusters at the beginning of 2006 to make it available for the community.

What's the difference between provisioning virtual and physical servers from OpenQRM's point of view?

The only difference is that provisioning selects a different resource-type for deployment. The deployment-mechanisms are exactly the same for virtual machines and physical servers. e.g. in openQRM the system-administrator sets up the data-center services using the VE (Virtual Environment) abstraction. A VE-configuration includes parameters which defines "how" a service should be deployed, which boot- and filesystem-image it should use, if it should be a cluster, how many nodes it should get, CPU- and memory requirements, high-availability etc. It also includes a parameter which sets the "resource-type" for a VE. The "resource-type" in openQRM can be a physical system or a virtual machine from different types. Virtual "resource-types" for VMware, Xen, QEMU and Linux-VServer are available and implemented as addtional plugins for openQRM. This means a system-administrator may decide at any time to move a service (VE) from a physical system to a VMware-partition, from there to a Xen-partition, next day running it on QEMU, then on Linux-VServer or back to a physical server by just changing the "resource-type" of a VE. This conformation of the different virtualization technologies and its unification with the deployment of physical systems enables the system-administrator to easily scale service up and down, provide instant-high-availability and to find and move to the virtualization technology which fits best to the service needs and requirements.

Is the rapid evolution of the virtualisation landscape a good thing for OpenQRM?

Yes, definitely

Can you give an concrete example where OpenQRM has been a life saviour?

Common situations where openQRM is a real benefit are e.g.:

  • Web-hosting companies which are using openQRM with Linux-Vserver and/or Xen virtualisation to host root-server for their customers. With openQRM's automated resource-planning and rapid, appliance-based provisioning those web-hoster are able to save about more than 50% of their deployment, administration and maintenance time and costs.
  • QA departments are usually benefiting most of openQRM's rapid deployment features and the capability to snapshot and clone specific server-images for their product testing. It helps them to increase the quality of their own product and decrease the testing-time by having a flexible server environment with lots of automatism.
  • Users of data-base service like openQRM very much because of its enhanced monitoring features via the Nagios-plugin and it's instant high-availability using hardware- and/or application-fail-over automatically triggered by openQRM. Those users reporting that database-downtime has decreased rapidly when they moved to managing their services with openQRM.

 

What can we expect in the future from OpenQRM?

The goal is to make openQRM "the" standard for modern data-centers. We are working on this target and you are more than welcome to join our effort.