Business Situation
Solution
Based on Microsoft Windows Server® 2008, Hyper-V™, System Center Virtual Machine Manager 2008 and System Center Operations Manager 2007, Avanade, Microsoft® and Accenture® developed an infrastructure solution, covering all of Deutsche Telekom’s requirements. Virtual Machine Manager includes a library enabling the saving and archiving of server and hardware configuration templates, configuration scripts, software images - combined with Deutsche Telekom's innovative IPTV and media services - and even complete server images. Intelligent positioning of virtual machines ensures optimum use of hardware resources. For this reason, virtual machines are distributed to as few physical systems as possible. This improves the energy balances and contributes to "Green IT". An integrated monitoring system provides early risk recognition of critical system states caused by high processor or memory loads, or slow response times. Fully automated, predefined actions can be performed independently due to its full integration with the Virtual Machine Management - shifting virtual machines to other physical hosts or starting additional virtual machines. Thus, responses are based on actual load behaviour of virtual machines at runtime.
Results
Avanade, Microsoft and Accenture introduced a virtual test and development environment at Deutsche Telekom's Products & Innovation division with run-time optimization and lifecycle management. Deutsche Telekom is now able to significantly shorten the time to market for technology infrastructure delivery. The implemented solution achieves optimum dynamic load distribution and best performance of virtual and physical systems at runtime. Intelligent positioning of virtual machines ensures optimum use of hardware resources. For this reason, virtual machines are distributed to as few physical systems as possible. This improves the energy balances and contributes to 'Green IT'. The new monitoring system analyses the entire test environment and prevents system downtime. Peak loads are immediately recognised and managed by automated responses, including shifting virtual machines to different physical machines or providing additional virtual machine.