* Have you ever considered infrastructure virtualization?
You’ve implemented server virtualization, and maybe even desktop virtualization. Have you ever considered infrastructure virtualization? Read on to learn about a software solution that allows enterprise data centers to dynamically change the workloads that various servers are running as well as which network and storage topologies those servers can access – without needing to make physical machine, cable, LAN or SAN changes.
In a typical data center, it’s common to provision more servers and storage devices than are needed in order to handle forecasted peak workloads as well as high availability and disaster recovery needs. This results in vast underutilization of expensive resources. In fact, server utilization rates are often less than 25%, which is why server virtualization has caught on like wildfire in the past few years.
How would you like to go beyond virtualizing a single server and its applications? What if you could control all the servers in your data center – physical or virtual – and repurpose them on the fly based on specific business or technical triggers? If this piques your interest, then get to know Scalent Systems and its Virtual Operating Environment (V/OE).
Scalent V/OE is a software solution that allows enterprise data centers to dynamically change the workloads various servers are running as well as which network and storage topologies those servers can access without needing to make physical machine, cable, LAN or SAN changes.
But Scalent V/OE isn’t an operating system like VMware. Rather, V/OE is a control system that provides real-time management and automation across physical and virtual servers. The Scalent controller (server) software runs on one machine, off to the side of your data center. Scalent V/OE reaches out to each physical or virtual machine to “tap” them via lights-out management protocols and cause them to power on / off. Then as a machine boots, Scalent – in real time – sets up the right network and storage addressing.
For example, if the machine is about to become an e-mail server, Scalent attaches it to the e-mail network and storage. If the machine is destined (at this particular moment) to be a VMware ESX server, Scalent attaches it to the correct ESX cluster network and storage for that cluster. Scalent then causes the machine to network-boot the appropriate bare-metal software image that the machine can now find on the central storage to which it has just been attached. Notably, that bare-metal software image can be Windows (and an Exchange server, or SQL server, or whatever), or Linux (and Apache, SAP, Oracle or whatever), or Xen or VMware ESX (and its cargo of virtual machines).
Of particular interest to data center operators are both the image portability and the fact that Scalent stays out of the data path. With image portability, Scalent can boot the server image into a physical machine or a virtual machine, with no “processing,” in real-time. Your Microsoft Exchange e-mail server can boot in a VMware virtual machine this boot, and the next boot, if you want, it could be on a bare-metal Dell blade, and the next on an HP rack-mount server. It’s all transparent.
“Out of the data path” means that Scalent gets out of the way once the machine starts booting. No data flowing to and from the machine passes through the Scalent controller, so no overhead is added. What’s more, if Scalent failed, no server is brought down; you just revert to your static data center.
By leveraging the functionality of VMware 3.5 and integrating Scalent software with both the network and SAN, V/OE creates portable instances that are bootable from the SAN, therein enabling a single boot image to be cloned and deployed to one or many servers, either physical or virtual. Once deployed, a server also can be migrated automatically in case of failure.
In essence, Scalent V/OE allows physical and virtual server repurposing to become nearly instantaneous. When employed, organizations can rack once, cable once, and reconfigure as needed without having to touch cables or machines. V/OE does this by disassociating connections from physical linkages. Your servers maintain their same addressing and connectivity even as they are moved, and boot from the SAN.
Scalent V/OE is used most often able to implement cost-effective disaster recovery, IT consolidation and development/test operational solutions. For instance, in the case of business continuity, Scalent eliminates the need for having a stand-by server for each of your production servers. If a production server fails, its applications can immediately be brought up on another server. Scalent allows an N-to-1 configuration for failover, lessening the cost of continuity planning. You can read some other use cases here.
Numerous technology companies – including EMC, Dell, HP, Sun, VMware and many others – partner with Scalent to deliver joint solutions. For example, HP recommends that customers use Scalent V/OE with HP BladeSystem c-Class servers in order to reduce costs, increase flexibility and enhance performance.
Read more about Scalent solutions in this Network World article on Tech Start-ups That Should Matter.
