7 ways Windows Server 2012 pays for itself
Windows Server 2012 is a monumental release packed with new features that touch every facet of the operating system. You'll see changes ranging from how data is stored on disk to the protocol for moving data between client and server and much more in between. The major design themes of the new server OS, which center on continuous availability, reduced cost, and lower management overhead, show up in many ways.
One of the principal architects on the Windows Server 2012 team is Jeffery Snover, who made an observation about this version of Windows Server that's worth repeating: "Microsoft has traditionally taken a few versions of major software projects to get them to the point of full maturity. It's significant that Windows Server 2012 has quite a few version 3.0 pieces including Hyper-V, PowerShell, SMB, and more."
Sure enough, these 3.0 charms alone are worth the price of admission, but a few newer features further sweeten Microsoft's server offering. In fact, I've identified seven Windows Server 2012 features you might call "supersavers." Listed in order of impact below, these features commodify high-end functionality, eliminate the need to purchase third-party software, reduce OS care and feeding, or in the case of PowerShell, offer the potential to save vast numbers of man-hours.
Any one of these features could make Windows Server 2012 a compelling upgrade for you. Perhaps the best part is they're all available in the Standard edition.
Windows Server 2012 supersaver No. 1: Storage Spaces
One of the main themes of Windows Server 2012 is the resiliency of all resources. For disk-related resources, the two new features are the Resilient File System (ReFS) and Storage Spaces. ReFS is the heir apparent to the venerable NTFS originally introduced with the release of Windows NT 3.1 in 1993. NTFS has obviously stood the test of time over the last 19 years with untold numbers of systems still using it today. Windows Server 2012 continues to support NTFS and undoubtedly will for years to come.
ReFS changes the way data gets written to disk. NTFS was susceptible to corruption of the file metadata -- the information the operating system uses to retrieve a file. ReFS uses an allocate-on-write method whenever any updates occur to prevent in-place corruption issues. It also uses checksums for metadata as another measure of validating saved data; you have the ability to enable checksums for the data as well. Microsoft calls this use of checksums Integrity Streams. It's a way to provide a measure of file protection even when the underlying disk system does not.
Storage Spaces has the potential of saving you significant dollars over a typical RAID-based disk array. That's because Storage Spaces works with raw disk drives arranged using JBOD, or "just a bunch of disks." Storage Spaces doesn't require any special (meaning expensive) disk controller unless you're building a cluster. Physical storage is allocated to a storage pool from which virtual disks, or spaces, get created. Virtual disks are, in turn, formatted with either NTFS or ReFS.
When you create a storage volume, Storage Spaces offers three different layout options -- simple, mirror, and parity -- that roughly equate to RAID 0, 1, and 5, although the algorithms used for distributing the data are totally different. Storage Spaces also provides the ability to "thin provision" volumes, which means you can create volumes of a virtual size larger than what is actually available in terms of physical capacity. More physical storage can be added to the pool to increase the physical capacity without affecting the virtual volume. This ability to add storage without incurring downtime is obviously a significant advantage when high-availability applications are involved.
The venerable CHKDSK utility is a major beneficiary of file system improvements. A new disk corruption scanner runs in the background on NTFS volumes, identifying correctible errors and data corruption. Most data corruption issues can be handled without the need to reboot the system and run CHKDSK to repair. If CHKDSK does become necessary, it can complete all operations in a matter of seconds -- versus the many minutes or even hours, in the case of large RAID disks, that it takes in previous Windows Server versions.
There are a few gotchas with ReFS, though none are showstoppers. You can't boot from a disk formatted with ReFS, nor is ReFS supported for removable media. More significant, you cannot convert an NTFS volume to ReFS in place, meaning you must copy the data from an NTFS volume to an ReFS volume.