ProSync CIO Tori Boats
When you ask for a customer’s trust, credibility goes a long way when you can show that you practice what you preach. We started with virtual desktops, virtualizing a demo system and some outlying workstations. It went so well, we went full board and virtualized all workstations at ProSync HQ. Now when Mike Lessing and the sales team bring customers through, they can see that we use virtualization for almost all of our business functions.
There was a need to push virtualization even further. I wanted to offer customers an environment where they could physically touch our virtualized solution. It isn’t vapor-ware hokus pokus, Dear Customer, it’s real. After speaking with our virtualization experts performing in-production virtualization for our customers, I was confident virtualization was ready to hit our back-end. So off we went.
What is Server Virtualization?
Very basically, it is running a server function, such as authentication and authorization services (AD, NIS, LDAP) within its normal operating environment (OS and application stack) with one physical change. While the logical stack remains the same: Hardware devices –> OS –> Application, then the physical stack changes ever so slightly. A hyper-visor is introduced. Simply, a software item that sits on top of the hardware and under the OS to manage hardware resources. The hardware appears as native and physical to the OS –> Application, but the hardware is “shared” among several OS –> Application groups (or commonly known in the physical world as hosts / servers).
Many benefits exist to Server Virtualization, but I’ll list a few major ones that we used for justifying this migration;
1. Higher Utilization of Server-Class Hardware:
Instead of one OS and Application stack using 10-30% of the hardware resources such as memory, CPU power, and network connection. Virtualization allows many OS and Application “Servers” (think of it as hosts with a server name and IP) to run on ONE hardware server (such as a single Dell 2950) using 80% of the hardware resources.
2. Fault-Tolerance, Availability and Disaster Recovery:
Backing up a server install is so difficult (not the data, but the OS, configuration, and installed applications, add-ons, etc). There is Ghost and Altiris and a myriad of ways to do this, but integrity of the backup and current-ness of the backup are always in question.
3. Restoration is Difficult. Did You Remember to Backup the System State???:
Virtualization allows not only a clone (think UNIX dd, but exponentially better) of the entire system (OS, Config, Apps, and Data), but an approx-instant restore and even running 2 or more of the same “servers” at once in different places.
Virtualization allows the “server” to be resource balanced on different server hardware, allows over-clocking of the
Desktop Virtualization
hardware resources, has a built-in fault tolerance (manual or automated)! The fault tolerance scenario: the server hardware goes down, the virtualized server (server OS, configuration, and applications, data all encapsulated into a little package, if you will) gets booted up on one of the other hardware servers. Bam, back in business!
The list goes on and also offers amazing “template” workstation abilities that go to the desktop virtualization and on-demand server processing power needs, but let’s stop there.
This has a significant impact to ProSync’s disaster recovery scenarios and availability. Not to mention customer credibility and skill building opportunities. Check it out!
