With the launch of Microsoft’s Windows Server 2012, enterprises should look for better management and an option that can squeeze more performance out of every physical server, according to an IT certification expert.
Server 2012, which became generally available this week, enables automation for certain server tasks that can cut the time it takes to configure servers by automating the process and save money by freeing up time for the most skilled admins to do other work,
Best Microsoft MCTS Certification, Microsoft MCITP Training at certkingdom.com
The higher skilled workers can now create automations via Power Shell, Jones says, leaving them with more time free to accomplish strategic tasks. At the same time, automating processes makes it possible for lower level admins to oversee automated tasks. That gets the tasks done faster and also reduces errors that can be introduced when these chores are done manually.
“It makes them more productive and less dangerous,” he says.
Some of this management could entail clicking a single button on 50 different servers, Jones says, something that can be done centrally in much less time.
Also in Server 2012, admins can knock off the graphical user interface from server software, turning devices into simple servers, Jones says. Tools for setting up Active Directory or DHCP servers, for example, would reside on a client, not the server itself.
Lopping the GUI off the server reduces the load on the server CPU, he says, which in some cases can be a significant reduction, citing an extreme case with an actual customer in which the business was able to squeeze 40% more virtual servers on a single host. “It’s a lot of overhead. I think that’s what most folks don’t realize,” he says.
Removing the GUI requires a single uninstall command and reboot. Jones recommends doing this on a virtual machine and testing whether the server works well with the client-side interface. He says the preferred platform for System Center and SQL Server is a GUI-less server.
Server 2012 also has support for blending in resources from traditional data centers, private clouds and public clouds, but these are options. If they aren’t useful to a particular business, they don’t have to be used, he says.
“All businesses will do it eventually,” for at least some servers, Jones predicts, “unless there is a major, major reason to hold all resources in-house.”