• E-mail this page
• |  Print this page
• | 
• | 
• | 

Infrastructure Virtualization

Server virtualization helped justify and broaden the use of the SAN by leveraging networked storage to enable features like server motion. In similar fashion, companies such as Scalent Systems are using infrastructure virtualization to further justify and broaden the use of a SAN by bringing those server virtualization capabilities to nonvirtualized systems: the ability to move or start new application instances in a matter of minutes after powering on and booting what was a cold, bare-metal server.

To bring this capability to traditional physical servers, several limitations needed to be overcome. The first limitation was the operating system itself. Prior to the server virtualization, once you created a database server it was always going to be a database server until you decided to reformat it and start over, which, of course, is very time consuming. Server virtualization tools solved this but created a limitation of their own: the hypervisor. With the hypervisor comes a performance penalty, and you still don’t have the ability to power down components. To move virtual server instances requires another virtualization host be powered on and running. Finally, there are some servers in the data center that you are just not going to virtualize.

With infrastructure virtualization you can power on a server, boot a raw image (OS and application) from the SAN, and take no performance hit. This allows you to create a small group of physical servers that are powered off but connected to the storage and IP networks. If you need to perform tests on that database environment, simply power on one of the servers from your group of idle servers, boot the correct OS image and application stack from the SAN, and begin your tests at full speed.

This works for virtual server hosts as well. If your computer power needs peak for some reason and you need to add more computer power to your server virtualization farm, you can power on a server from that idle server group and have it boot the virtualization operating system. Once booted, use your virtualization console to begin reallocating virtual servers on to it. Infrastructure virtualization is not a competitor to server virtualization; in fact it is a complement to it.

Of equal importance is when the test is complete or the peak need in the virtualization farm is finished, those servers can be returned to the idle pool in their bare-metal, powered-off state.

The second limitation is that the servers (physical and virtual) have an IP address, and other users, servers, and network resources are expecting to see that server at that address. If you need to manually reconfigure these resources before or after moving and activating the new server, the fluidity of the process is lost. An infrastructure virtualization environment moves the LAN context with the server. All the network resources are reconfigured automatically.

The final limitation is the storage network. Host bus adapters and even virtual HBAs are hardbound to certain LUN segments on the storage network. If you move a server you could be faced with reconfiguring potentially hundreds of storage access permissions or, alternatively, configuring every physical machine to have access to every storage resource, which can create some significant security and management implications. With infrastructure virtualization, the storage network context moves with the server’s new physical location.

Server virtualization hosts suffer from the same IP and storage network limitations mentioned above. Infrastructure eliminates the need to create flat open networks that are the common workaround to the internetworking issues created by virtualization, as well as provide the ability to perform as needed to scale out of the virtual farm when peak loads occur.

With infrastructure virtualization, you can increase the return on your SAN investment further than ever as well as further reduce physical server counts and lower power consumption.

George Crump is founder of Storage Switzerland, an analyst firm focused on the virtualization and storage marketplaces. It provides strategic consulting and analysis to storage users, suppliers, and integrators. An industry veteran of more than 25 years, Crump has held engineering and sales positions at various IT industry manufacturers and integrators. Prior to Storage Switzerland, he was CTO at one of the nation's largest integrators.

« Business Tech Is The Land Of The Living | Main | Joomla Rocks, Or How To Build A Professional Web Site For No Money Down »