File services are worth a second look

Today, I want to look at the new file services that will come with Windows Server 2003.

Wait! Get your pointer off that delete button!  There really is something exciting to look at in file services.

For those of us who have dealt with Microsoft and file systems since the days of DOS 1.0 it’s hard to imagine that very much new and exciting could occur. In those days, the concept of “directories” (what we now call “folders”) still didn’t exist – everything was kept in what would become the root directory of a disk. Of course, a disk only held 128K bytes of data, so it wasn’t too confusing. But a look at the list of new (or improved) features intended for Windows Server 2003’s file system shows the following:

* Virtual Disk Service.

* Volume Shadow Copy Service.

* Distributed File System.

* Shadow Copy Restore.

* Offline Files Improvements.

* WebDAV Redirector.

* Web User Interface for Server Administration.

* Command-line Tools.

* Automated System Recovery.

* Chkdsk.

* Disk Defragmenter.

Chkdsk, for example, certainly isn’t new. In fact, with Windows NTFS journaled file system (which tracks file changes in transactions rather than the traditional linked list used in the FAT system), CHKDSK is rarely needed. Its strength is in recovering disk sectors used in files but that are no longer linked to a file. Microsoft estimates that less than 1% of server crashes (which also are much less frequent in Win2K3) will be necessary to run CHKDSK. Nevertheless, the utility has been improved so that it performs between 20% and 40% faster than in Windows 2000.

Brand new features include:

* Virtual Disk Service, an abstraction layer that allows identical access to disk media no matter where it resides – SCSI or IDE drives in the physical server.

* USB connected storage devices.

* RAID systems.

* Network-attached storage devices plugged into an Ethernet hub or switch.

* Storage-area network devices using Fiber Channel or iSCSI connections.

A boon to network managers is the new Volume Shadow Copy Service. A shadow copy of a storage volume is a point-in-time copy of the original entity. The shadow copy is typically used by a backup application to back up files that are made to appear static, even though they are really changing. And if created on a SAN the shadow copy can be transported to another server for backup, testing, or data mining. Coupled with the clustering services we talked about last week, data loss through hardware or software failure should soon almost entirely disappear from your network – provided you understand and take full advantage of these new features.

There’s a whole lot more available in new and improved file services, so come back next time and we’ll take another look.