Windows Server 2003 Gets a Few Tweaks

Gearing up for the April 24 release, Microsoft fine-tunes technology, programs and services for Windows Server 2003.

Microsoft Corp., trying to make up for the disappointing uptake of Windows Server 2000 and that softwares headache-inducing Active Directory installation process, is expected to announce today tweaks to technology, programs and services for Windows Server 2003.

Bob Ellsworth, director of Microsofts Windows Server Product Management Group, in Seattle, said that a key motivation for the reliability innovations is to convince Windows NT 4 customers to jump over Windows 2000—which has had a disappointing record of customer uptake—and head straight for Windows Server 2003. "The messaging were using is, Boy, if youre on NT 4 and you havent yet moved to Windows 2000, nows the time," Ellsworth said. "Its ready for prime time."

Part of being ready for prime time will include wizards to help implement Active Directory, which many enterprises ignored, due to its laborious install, Ellsworth said.

Tweaks to services around Windows Server 2003, which is scheduled for release on April 24, include a new version of the companys Datacenter High Availability Program. The program, originally introduced in September 2000, is geared toward satisfying the needs of Microsofts enterprise glasshouse customers—those that rely on 24x7 uptime in their data centers.

The new program, the Datacenter High Availability Program for Windows Server 2003 Datacenter Edition and Windows 2000 Datacenter Server Customers, draws on experience gleaned from participation of Microsofts partners and customers and will feature three new enhancements.

The first will be certified configurations provided through partner OEMs, aka the EQP (Enterprise Quality Program). EQP is geared toward ensuring that customers get complete, pre-tested hardware and software configurations.

Microsoft has streamlined the programs testing process, which used to entail OEMs getting through 14 days of hardware compatibility without failures. If any component failed, the entire configuration would have to be retested. Under the new program, OEMs will test the total configuration only once. If a component fails, only a one-day retest of the entire configuration will be required.

Ellsworth said that the streamlining is expected to reduce the cost of retesting and will get reliable configurations to customers faster.

Datacenter High Availability Program Customers also will be able to support any Gold Certified Datacenter Service Provider—including OEMs, resellers, systems integrators and from Microsoft directly—as opposed to being limited to support from OEMs.

Finally, Microsoft added a new support service to the Datacenter High Availability Program, called High Availability Resolution Queue, that will include 24x7 global support.

Microsoft is also introducing a slew of new features to enhance product reliability. They include Windows driver protection, a feature that prevents problematic drivers from installing and directs customers to an updated version. Also, a kernel mode driver blocking feature is designed to give administrators fine-grained control of driver installation on the server. Driver rollback is designed to get Windows Server 2003 back up and working more quickly when faulty drivers affect performance.

Other reliability features include process isolation, a new fault-tolerant architecture that isolates applications from one another and from the servers kernel; application recycling for IIS 6, wherein new instances of applications are restarted side-by-side to avoid failure due to memory leak and fragmentation; and fault tolerance and new clustering capabilities that streamline setup and monitoring of cluster nodes.

Finally, Microsoft will announce the MRS (Microsoft Reliability Service). This service will monitor event data from customer data center servers, analyze the data, and produce customer-tailored reliability and availability reports. Officials said that MRS will enable customers to monitor availability of software and hardware components within their systems, identify root causes of planned and unplanned downtime, and take preventive measures to improve system availability.

Read more articles by Lisa Vaas
Read more articles on Microsoft
For more Microsoft scoops, check out Ziff Davis Microsoft Watch
For more on Windows Server 2003, see our special section.