Export (0) Print
Expand All

Optimizing Branch Office Servers by Using Windows Server 2008 R2 with Hyper-V

Technical Case Study

Published: September 2011

Microsoft Information Technology (IT) used Windows Server 2008 R2 with Hyper-V to virtualize all of the enterprise's branch-office servers. This enabled Microsoft IT to extend virtualization's benefits to a variety of site types, from small offices to large, complex branches, thereby improving services, reducing costs, and expanding infrastructure-consolidation efforts.

Download

Download Technical Case Study, 243 KB, Microsoft Word file

Customer Profile

Situation

Solution

Benefits

Products & Technologies

The vision of Microsoft Corporation, as the worldwide leader in software for business and personal computing, is to enable people and businesses throughout the world to realize their full potential. Microsoft employs 89,000 people and had revenues exceeding $62 billion for the year ending June 2010.

The Microsoft IT branch-office strategy is to streamline communications and increase operational efficiency. With 223 branch-office servers running outdated operating systems and nearing the end of warranties, Microsoft IT replaced them with Virtual Branch Office Servers (VBOS) that offer extended services.

Microsoft IT used Windows Server 2008 Hyper-V technology to virtualize 223 branch-office servers. This provided a secure, future-proof platform on which to provide key services to users.

  • Increased the IntelliMirror user base and maintained low total cost of ownership. The average monthly cost for each IntelliMirror user is down to approximately $2 U.S. dollars, which is an excellent return on investment.
  • Centralized administration and support by configuring Integrated Lights Out features, which enable remote-administration access to VBOS
  • Reduced hardware and support savings significantly (approximately $120,000 USD) by replacing 28 physical domain controllers with Read-Only Domain Controllers.
  • Windows Server 2008 R2 SP1
  • Microsoft System Center Virtual Machine Manager 2008 R2 SP1
  • Microsoft System Center Data Protection Manager 2007 SP1
  • Remote Server Administration Tools for Windows 7 SP1
  • Hyper-V
  • Read-Only Domain Controller
  • File Server Resource Manager
  • BranchCache bandwidth optimization
  • Folder Redirection
  • Server Message Block
  • Microsoft Operations Framework 4.0
  • Windows PowerShell

Situation

Microsoft has a highly dynamic IT environment, where released-to-market software exists alongside the next generation of Microsoft products that are at various stages of their release cycles. Therefore, Microsoft IT faces a constant challenge when trying to align with the business' rhythm, while simultaneously testing and adopting new software.

Server virtualization is one area in which MSIT's drive toward early adoption has influenced product development significantly. Since 2005, Microsoft IT has pursued virtualization aggressively in its data-center environment to reduce costs and minimize server sprawl. Server utilization data from Microsoft® System Center Operations Manager 2007 was crucial to this virtualization drive.

Microsoft IT spent several years analyzing utilization metrics for central processing unit (CPU) servers. This data indicated that the average production server's CPU used less than 7 percent of its capacity. Additionally, the data showed that this figure was falling consistently as processor speeds improved. Therefore, Microsoft IT determined that the best way to improve CPU usage was to consolidate and pursue virtualization.

The Microsoft IT branch strategy is to streamline communications and increase operational efficiency on a secure, future-proof platform by utilizing the Windows Server® 2008 Hyper-V™ technology. The success of the data-center virtualization paved the way for Microsoft IT to pursue virtualization of its branch office servers, which it began in mid-2008. During the one-year initial pilot project, Microsoft IT used Hyper-V to deploy Virtual Branch Office Servers (VBOS) to replace 91 older, end-of-warranty servers, many of which could not run newer versions of the Windows operating system. Microsoft IT then replaced 132 servers in the second phase of the VBOS deployment.

Solution

As part of its role, Microsoft IT supplies services to more than 190,000 users in approximately 600 buildings worldwide. Microsoft IT guarantees the best possible user experience by managing all IT services to the Microsoft Operations Framework 4.0 guidelines, and offers 99.99 percent availability targets on all service-level agreements (SLAs). Microsoft IT tracks its service levels on detailed business scorecards that have defined minimum performance levels and apply to services running on prereleased software, as well.

Microsoft IT then develops Service Improvement Programs (SIP's) based on these scorecards, to ensure that it consistently meets, and improves upon, SLA targets. Microsoft IT does this by collaborating with business units and corporate development teams to ensure that the requirement to validate new products does not compromise service delivery targets. This partnership enables Microsoft to develop products that help the company maintain its leading position in enterprise-software solutions, while continuing to deliver a stable IT infrastructure for its users.

Changing the Infrastructure

Deploying VBOS servers was the biggest change to Microsoft's branch-office infrastructure since the 2005 Model Enterprise Initiative (MEI) project. This project created a largely centralized infrastructure model where regional data centers deliver core services, such as Microsoft® Lync™ and Microsoft® Exchange, and a network of branch-office servers supplies a smaller subset of services.

As part of MEI, all Microsoft branches were classified against a set of standards that determined what infrastructure each site required. When planning the placement of branch-office servers, Microsoft took into account factors such as site location, local business requirements, the connectivity options available, and the site's function in the organization. This resulted in a standard method of determining which offices had branch-office servers and which had no server.

VBOS is an internal Microsoft IT specification that details the configuration of a physical server that hosts multiple services. In general, having a VBOS server at a remote field office lessens the site's server load and its dependence on an expensive wide area network (WAN) link. Furthermore, service consolidation reduces costs through lower hardware, energy, and management overheads, and creates a more dynamic IT infrastructure. Additionally, it ensures that branch-office users have a common set of standard services available locally.

Improving Branch Offices by Deploying the First Stage of VBOS

Prior to the initial VBOS deployment, the branch-office server design was a single server running Windows Server® 2003 with Service Pack 2 (SP2) and six core services, including:

  • File

  • Print

  • Data Distribution Services (DDS)

  • Windows® Deployment Services

  • Microsoft System Center Configuration Manager 2007 (SCCM)

  • Microsoft IntelliMirror® management technologies

Microsoft IT determined that these services needed to be located on local servers so it could meet minimum performance levels specified in an SLA. However, because this was a functional system, with multiple services sharing the same operating system, there were a number of security and performance concerns, including:

  • Interdependencies between services. The likelihood was high that updates or upgrades to one service would affect other services. This, in turn, would limit the ability of service managers to develop their services independently or to guarantee the 99.99 percent SLA targets.

  • Addition of new services. The addition of new services into this environment was considered high risk, which in turn meant that the platform was extremely limited.

  • Concerns about security. The ability of administrators to have blanket access to services meant someone might inadvertently affect the availability of services.

Therefore, the VBOS refresh program aimed to improve the branch office experience by:

  • Improving security on the branch-office platform and individual services.

  • Improving service management by isolating services into separate virtual machines.

  • Using virtualization to extend the number of services available to branch offices.

  • Leveraging technology advances to improve WAN performance.

  • Bringing all branches equal in relation to a minimum set of hardware standards.

The Initial VBOS Pilot Deployment

Microsoft IT tried to resolve these issues with the initial deployment of virtual servers into branch offices by using the Release to Manufacturing (RTM) version of a Windows Server 2008 Server Core for the host operating system, with four virtual guest servers.

The plan was to deploy the existing six services in a more secure way, making them less dependent upon other services, and by extension, enable service managers to develop and improve services. Initially, the File and DDS services shared a virtual machine, as did the IntelliMirror and Print services, and Windows Deployment Services and SCCM were moved to standalone virtual machines.

For the pilot deployment's first stage, Microsoft IT developed stock-keeping units (SKUs) for small branches (no more than 400 users) and medium branches (no more than 1,000 users). Meanwhile, the large and High Availability branches were deferred to the second wave to minimize the business impact. All deployments relied on IT Managers racking the servers and running a Windows PowerShell™ script that built the standard configuration on each new VBOS server. The one script worked for both the small or medium configurations.

This initial VBOS deployment was a big success, and produced a simple and repeatable process that enabled Microsoft IT to build and deploy a fully commissioned VBOS server with minimal training requirements. Additionally, there were several benefits to the project, including improved security and more effective service management. Therefore, Microsoft IT decided to replace its remaining 130 branch-office servers with VBOS servers. The clustered design for large sites enabled Microsoft IT to add the nine large and high availability sites to the branch-office server program for the first time.

Lessons Leveraged from the Initial VBOS Deployment to the Second Phase

Microsoft IT designed the initial VBOS pilot to allow allowed variances between sites as a way to test different scenarios. This, in turn, enabled Microsoft IT to streamline its approach in the second stage of deployment. The lessons that Microsoft IT carried with them to the second phase of deployment centered on design, deployment, and security, including:

  • The SKU design that Microsoft IT used for the pilot branches enabled the six original services to be delivered on the more secure and efficient Hyper-V platform. This worked well, although the decision to congregate certain services--DDS and File, and IntelliMirror and Print--on specific virtual machines essentially bound those services together. However, this caused issues as service managers attempted to manage their services independently. Microsoft IT opted to change this approach in the second deployment phase so that each service had its own virtual machine, which ensured a cleaner deployment of services, and enabled managers to maintain control of their specific services.

  • Microsoft IT enabled upgrading between small and medium SKUs, as well as adding extra storage to the SKUs to extend the IntelliMirror and File Share virtual machines. Further, Microsoft IT retained the pilot's upgrade path for the second deployment. However, the additional storage for individual services added a layer of complexity that the VBOS deployment team wanted to avoid in the second phase.

  • Local IT managers were responsible for racking, building, and onboarding the servers during the first deployment phase, and used a Windows PowerShell script to build the standard configuration. This guaranteed that no build variations would occur between sites. However, there were instances where the script did not run correctly, which caused delays if the site's IT manager could not resolve the issue.

During the second deployment, Microsoft IT centralized this process because many IT managers are remote resources that control several sites spanning a large geographical area. Therefore, Microsoft IT used the internal IT Service Management (ITSM) tool to centralize and simplify the build and support process. ITSM is a central database that controls every aspect of a server's lifecycle from build to decommission.

  • The VBOS deployment team wanted to restrict server access to the central engineering and build teams who managed the service deliverables and SLA measurements. In the first deployment phase, the deployment team chose a Server Core installation of Windows Server 2008 as the host operating system, which decreased the number of updates and subsequent reboots significantly.

Additionally, segregating services into separate virtual machines limited access to individual services. However, the VBOS deployment team felt that further restrictions, to the central engineering and build teams only, would release IT managers to manage service delivery and improvement projects.

Designing the Second Phase of the VBOS Deployment

The VBOS deployment team improved upon the initial rollout by leveraging lessons learned from the initial deployment and producing a standard platform for all remaining branch offices, including large and high availability sites. The plan was to deploy the original six services in separate virtual machines, with a hosted branch cache virtual machine on each server and read-only domain controllers (RODCs) in branches deemed appropriated by the Active Directory® service team.

These services were running in field sites currently, so available historical data specified the resources that each service would require. Additionally, the VBOS deployment team consulted with each service manager regarding the amount of compute and storage resources they required for growth over a three-year period.

The deployment team then developed small, medium, large, and extra-large SKUs based on the cost model in the following table:

Table 1. SKUs developed for the second phase of VBOS deployment

Max. number of users

Storage Capacity

Shared Storage

High Ability

Hardware Cost per Virtual Machine (USD)

Small

440

3 terabytes

No

No

$1,625

Medium

1150

5.3 terabytes

No

No

$2,000

Large

2500

14 terabytes

Yes

Yes

$6,250

Extra Large

4500

19 terabytes

Yes

Yes

$7,500

The deployment team focused its design principles on high service availability and providing centralized support and management teams for VBOS servers, with the intention of standardizing the VBOS platform based on the Microsoft IT virtualization host offering that the Microsoft's data centers use, which included:

  • Using Windows Server 2008 R2 Core with Hyper-V

  • Using multiple network adapters configured as a team

  • Configuring hot spare-disk drives

  • Configuring virtual machines as highly available and with live migration enabled

The VBOS deployment team carried these same design decisions forward, and configured all servers to use Integrated Lights Out (ILO) features that enable full out-of-band management and remote-administration access to the branch office servers. This was pivotal to the centralized administration and support model that VBOS required, as many branch-office sites have no dedicated IT staff.

When considering hardware, the VBOS deployment team chose to use the same server model for each site type, so that branches had a natural upgrade path for adding additional storage. Additionally, the team opted to:

  • Configure SKUs for the physical host servers with 16 processor cores, 72 gigabytes (GB) of memory, and four network interface cards (NICs).

  • Provision each guest virtual machine with at least two, and no more than four, virtual processors.

  • Provision each guest virtual machine with at least four GB, and no more than eight GB, of memory.

  • Configure all virtual hard disk files are dynamic.

  • Configure the large and extra-large SKUs as two node failover clusters, connected to a shared store area network (SAN) storage array. For these sites, the team decided to configure storage as a single cluster shared volume and the virtual machines as highly available. This enables the live migration feature that Hyper-V includes that facilitates performance and agility.

The only component of guest servers that varies between sites is storage, making this a simple, easy-to-manage environment for the central support teams.

Deploying the Second Phase of VBOS

Microsoft IT made a number of changes to the VBOS deployment process as it embarked on the second phase, including:

  • With the exception of four countries, where customs processes and cost factors required the deployment team to purchase hardware locally, all the VBOS servers were bought and assembled in the U.S. before being shipped to specific sites, where the IT managers then racked them, and connected them to the network. The VBOS deployment team found that the bulk discount more than offset the freight charges. Additionally, it guaranteed that all servers were built to exact design specifications.

  • The nine large and extra-large sites required additional clustering and configuration work, due to the SAN implementation. The VBOS team also ordered, deployed, and configured these SAN units using central resources. Once the phase-two servers were racked, each server's build information was entered into a Change Request Form in the ITSM tool. The tool validated each change, and then allocated each to an engineering resource who ran the configuration script remotely. Ultimately, this resulted in each location having a common platform, and was much quicker and more manageable. In comparison, while more-complex branches were deployed in the second VBOS rollout, the build and onboard process was completed nearly 30 percent faster than during the first phase, which oversaw 91 VBOS servers.

  • The deployment team transferred ownership of individual virtual machines to server managers, and tasked them with full upgrade and updating responsibilities. Previously IT managers had been responsible for updating any servers that had somehow missed the monthly compliance deadlines. However, this change placed full responsibility for the performance of local services with the specific service manager who was accountable for the site's SLA.

Additionally, this enabled IT managers to concentrate on user management and other value-add projects, while the service managers managed service delivery. The VBOS deployment team, in partnership with field IT resources, felt that this more-secure option further reduced the number of resources with server logon access. Additionally, deployment times were reduced significantly because service managers were onboarding their own services remotely.

Improving Service in the Second Phase of Deployment

The upgrade to the branch-office servers offered a range of opportunities, beyond virtualization and management options. Allowing service managers complete control over a dedicated virtual server meant that Microsoft IT could adopt the benefits from future upgrades much faster, without affecting other services negatively. This approach also enabled service managers to leverage their new responsibilities to change their services for the better and often reduce costs.

The service managers for the SCCM, DDS, and File Share virtual machines all took advantage of the new opportunities, and reorganized their services to offer users a more streamlined, cost-effective service. For other services, such as IntelliMirror, domain controllers, and BranchCache®, the benefits were particularly impressive, as the service managers were able to deliver a range of new benefits after the second deployment.

Improvements to the IntelliMirror and Client Data Management Services

Service managers made significant improvements to data backup and recovery after the second phase of VBOS deployment. The Client Data Backup Management team offers several services that ensure the central management of user data, while providing users with fast, secure access to their data, independent of network connectivity. One key service that this team delivers is IntelliMirror.

The IntelliMirror service management team felt that the release of Windows Server 2008 R2 and Windows® 7, plus the rollout of Hyper-V in branch offices, offered sufficient additional functionality to justify a major service-improvement plan. Feature improvements in Server Message Blocks (SMBs), Folder Redirection, Offline Cache Folders, and Windows BitLocker® Drive Encryption all made service improvements readily available. This, in turn, improved the user experience significantly in the following areas:

  • SMBs: IntelliMirror uses SMBs to enable client computers to access files and services from the network's server programs. Windows Server 2008 and Windows 7 introduced a redesigned version of the protocol, known as SMB 2.0. This protocol is much more efficient at opening up a range of benefits that improve the user experience with the IntelliMirror product and improve network usage. These benefits include:

    • Larger read and write capabilities.

    • Better use of faster networks, even with high latency.

    • Improved message signing with changes to the Hash-based Message Authentication Code (HMAC). SHA-256 replaced MD5 as the hashing algorithm, which improved configuration and interoperability.

    • Request compounding, which enables sending multiple SMB 2.0 requests as a single network request.

    • Durable handles, which enable SMB 2.0 connections to reconnect transparently to the server if a temporary disconnection occurs.

  • File Server Resource Manager: File Server Resource Manager is a suite of tools that enable administrators to understand, control, and manage the quantity and type of data stored on computers that are running Windows Server. These advanced tools help administrators to monitor existing storage resources efficiently. Windows Server 2008 R2 supports File Server Resource Manager on all server installation options, including Server Core. However, at Microsoft, the IntelliMirror virtual machines sit on a full version of Windows Server 2008 R2 with the standard GUI.

  • Larger storage quotas: Hyper-V was a key enabler in the service-improvement plan to extend the range of user information that IntelliMirror helps secure. It enabled Microsoft IT to increase storage quotas for the service's users, and added desktop files and favorites to the range of files that IntelliMirror secures. This improves disaster-recovery capabilities. Additionally, because storage costs are lower on Hyper-V servers, Microsoft IT was able to increase minimum quotas from 1.25 GB per user to 3 GB, with the option for individuals to increase their quota, as needed, to 15 GB.

These benefits are significant, and are due to the IntelliMirror team leveraging the availability of VBOS to improve data security in the IntelliMirror product offerings and utilize BitLocker to secure offline cached copies of user data. This, in turn, ensured that data management met security and governance guidelines.

Additionally, there were cost benefits as the IntelliMirror user base increased, because users moved away from unsupported backup and recovery solutions. User enrollment in IntelliMirror has more than doubled since the VBOS deployment began, which is significant from a compliance and cost perspective. Over a span of 12 months, beginning in July 2009, Microsoft users generated costs of more than $1.4 million in U.S. dollars (USD) on various data-storage or recovery issues, in areas such as:

  • Costs for dispatch of support technicians to address data-backup and restore requests.

  • Costs for the time spent recovering data that was lost.

  • Costs for purchasing external hard drives for individual backups.

The improvements to the IntelliMirror service reduced this figure significantly by attracting users to the more-secure corporate solution. By capitalizing on the existing Microsoft worldwide enterprise-server infrastructure, Microsoft IT has increased the IntelliMirror user base while maintaining a low total cost of ownership (TCO). On average, the monthly cost for each IntelliMirror user at Microsoft is down to approximately $2 USD currently, which is an excellent return on investment (ROI). Prior to virtualization, the average monthly cost to provide the IntelliMirror service ranged from $3.50 to $5 USD per user, based on user counts and the cost to configure additional drive space, as necessary.

Benefits of Utilizing Read-Only Domain Controllers

Building RODCs into the design for the second VBOS deployment phase was a major step forward for Microsoft IT. RODCs first were introduced as new types of Active Directory Domain Services (AD DS) domain controllers in the Windows Serve 2008 operating system. RODCs offer significant benefits, including providing organizations with a more secure way to deploy domain controllers outside data centers and in less physically secure branch-office environments. Some key security features include:

  • Read-only database: The AD DS database on an RODC contains the same objects and attributes that a writable domain controller contains, with the exception of passwords. Unlike a full domain controller, users cannot change and replicate an RODC database back to the main Active Directory forest. This provides a new, additional layer of protection.

  • Unidirectional replication: RODCs have a read-only database, so users cannot initiate replication changes, and no domain controllers that are replication partners will pull changes from an RODC.

  • Role separation for administrators: In the Microsoft IT implementation of VBOS, the administration of server roles plays an important part in the overall security picture. While administrators perform hands-on maintenance work locally, only the data center's server management group has logon access to the VBOS host, and only the Active Directory team has logon access to the RODC virtual machine. Maintaining role separation enables Microsoft IT to offer an Active Directory service that provides security and protects the individuals that contribute to delivering the service's components.

The deployment of RODCs also resulted in cost and consolidation benefits. Virtualization improves upon the use of standalone servers as a branch's domain controller. Additionally, as new locations require domain controllers, or as existing domain controllers need replacing, Microsoft IT can leverage the existing VBOS host to provide a virtualized RODC. This is significantly more cost effective and efficient, and reduces server sprawl dramatically.

Building RODC resources into the SKU design for the VBOS deployment, and then providing the Active Directory team with the ability to utilize the VBOS platform, the deployment program paved the way for significant savings opportunities. In 2011, Microsoft IT replaced more than 28 physical domain controllers with RODCs on the VBOS platform. This reduced new hardware costs by more than $120,000 USD. Additionally, Microsoft IT was able to avoid spending money on the costs associated with installing, maintaining, upgrading, and supporting a physical server.

In comparison, the total cost associated with providing resources for an RODC on a VBOS server is 150 GB of disk space per server, plus shared CPU and random access memory (RAM) resources. This is significantly less than the cost of provisioning a new server and managing it during its lifecycle. Microsoft IT estimates that a typical server ultimately ends up costing three times its original cost, over a three-year period, and likely is underutilized, as well. Meanwhile, a virtual machine on a shared server costs much less over that same period, because it shares resources, and server utilization is significantly higher. Therefore, utilizing RODCs was an excellent ROI.

Implementing the New BranchCache Feature on Virtual Machines

The wide-scale implementation of hosted BranchCache virtual machines on the VBOS servers was a considerable step forward. An unintended consequence of the MEI consolidation project was an increased reliance on the WAN link, because applications and services were centralized. Implementing the new BranchCache feature, available in Windows 7 and Windows Server 2008 R2, provided Microsoft IT with a great opportunity to address this issue without additional costs.

BranchCache stores data locally in a branch office, so that when users in that office requests the same file, BranchCache retrieves it from the local cache instead of downloading it repeatedly over the WAN. However, for BranchCache to function successfully, client hardware must be running Windows 7 Ultimate or Enterprise editions, and the web and file servers must be running Windows Server 2008 R2. Additionally, administrators must enable the BranchCache feature.

Administrators can implement BranchCache in two different modes, both of which are in use at Microsoft:

  1. Distributed BranchCache: In a distributed scenario, BranchCache does not use a local server to host a cache, but rather enables cache storing on users' individual computers. This implementation is beneficial particularly in smaller offices where no local server is available. However, a drawback to distributed BranchCache is that multiple downloads of a file over the WAN do occur. When a user downloads a file, it thereby is saved to that user's computer. When that user is offline, and a second user tries to download the same file, a second copy is downloaded over the WAN.

  2. Hosted BranchCache: In a hosted implementation, BranchCache stores the cache on any branch-office server that is running Windows Server 2008 R2 and that has the host BranchCache setup. After the first download, other clients that require the same content retrieve it directly from the hosted cache. A major benefit of hosted BranchCache is that all clients in a branch office can access a single cache, even if they are on different subnets.

A core responsibility of Microsoft IT is to act as the enterprise's First and Best Customers, which means that Microsoft IT adopts early-stage products so that they can help product teams evaluate and improve them before release to market. Therefore, in keeping with this goal, Microsoft IT introduced a pilot to test hosted BranchCache on 14 of its VBOS phase-one deployment servers and 10 distributed BranchCache sites. The functionality that Microsoft IT tested during this pilot included:

  • Access to internal file shares via SMBs.

  • Access to an internal Microsoft Office SharePoint® Server 2007 sites by using HTTP.

  • Integration with System Center Configuration Manager and the ability to access advertised downloads by using Background Intelligent Transfer Service (BITS).

The results of these tests showed a 53 percent reduction in WAN usage across both pilots, when branch-office workers accessed locally cached data. BranchCache achieved these impressive results without requiring Microsoft IT to modify its security architecture or implement any new management technologies.

This was a particularly big win for branches with low-bandwidth challenges. Microsoft IT achieved the success of hosted BranchCache for the relatively small resource investment of 100 GB of hard-disk space per server, plus some shared CPU and RAM resources. As a result, Microsoft IT decided to install hosted BranchCache virtual machines across all VBOS servers deployed in the phase one and two rollouts.

Best Practices

Perhaps the biggest issue that Microsoft IT faced when embarking on the VBOS deployment was ensuring that the servers got to their final destinations in a timely and economical fashion. All 132 servers that Microsoft IT deployed during the second phase of the VBOS project were purchased in the United States, and delivered from Redmond, Wash. In the future, when new opportunities arise to deploy VBOS servers, Microsoft IT hopes to:

  • Determine what countries require new hardware, and then attempt to purchase it in that country. The hardware specifications may be hard to meet in a given country, but the potential savings, both in cost and time, could be significant. During the pilot project, one server spent more than a year working its way through its destination's customs process, and then Microsoft IT additionally had to pay a $2,000 USD release fee once it made it through customs.

  • Utilize documentation or other forms of communication to specify the processes and procedures to onboard virtual servers quickly and correctly. Some IT managers oversee multiple sites within a country, and the physical distance between locations can be significant. Documented, accessible onboarding procedures enable third parties to assist when an IT manager is unavailable in a timely fashion, and then Microsoft IT personnel in Redmond can get the servers up and running quickly.

Benefits

The deployment of VBOS at Microsoft branch offices around the globe resulted in several benefits, not the least of which was Microsoft IT ensuring that all offices were operating on the same platform. This makes support a much easier and cheaper task. Other benefits included:

  • A reduction in WAN usage by approximately 53 percent when branch-office workers accessed locally cached data from BranchCache virtual machines on VBOS servers.

  • Configuration of all VBOS servers to use ILO features, which enabled Microsoft IT to centralize administration and support. This is particularly important for branch-office sites that do not have dedicated IT staff, as it enables full out-of-band management and remote-administration access to the branch-office servers.

  • Significant cost savings, both in hardware and support. Building RODC resources into the SKU design for the VBOS deployment, and then providing the Active Directory team with the ability to utilize the VBOS platform, enabled Microsoft IT to reduce hardware costs by more than $120,000 USD by replacing approximately 28 physical domain controllers with RODCs on the VBOS platform. Additionally, Microsoft IT was able to avoid spending money on the costs associated with installing, maintaining, upgrading, and supporting physical servers.

Conclusion

The VBOS deployment program delivered, and in many areas exceeded, the core goals that Microsoft IT initially set. Overall, the Virtual Branch Office server program delivered a streamlined, efficient, user-friendly, and cost-effective service that has become even more valuable since its inception.

The VBOS deployment program opened up opportunities for the Service Management teams, by leveraging both current functionality and ensuring the capability to capitalize on future offerings. This enables service managers to continue developing their services and extend their service range to benefit their users. This increases service functionality in many service areas, including IntelliMirror and BranchCache. The current increase in service functionality, and reduced costs, has exceeded all expectations.

The deployment to the VBOS platform enabled Microsoft IT to leverage the wide range of benefits that virtualization provides to data centers to offer additional value-add to the branch-office environment. The project was a key opportunity for Microsoft IT to deploy a common scalable platform across all Microsoft branch types, where Service Managers could take advantage of the expansive new features available in the Windows 2008 R2 and Windows 7 operating systems.

VBOS has delivered a flexible, dynamic, and future-proofed platform to Microsoft's branch offices, thereby improving security, and each site's support model and range of available services. Additionally, the program has enabled Microsoft IT to consolidate servers and provide a more secure approach to service management, while simultaneously reducing the overall costs of deploying the branch-office infrastructure.

For More Information

For more information about Microsoft products or services, call the Microsoft Sales Information Center at (800) 426-9400. In Canada, call the Microsoft Canada information Centre at (800) 563-9048. Outside the 50 United States and Canada, please contact your local Microsoft subsidiary. To access information via the World Wide Web, go to:

http://www.microsoft.com

http://www.microsoft.com/technet/itshowcase

© 2011 Microsoft Corporation. All rights reserved.

Microsoft, BranchCache, Hyper-V, IntelliMirror, Microsoft Exchange, Microsoft Lync, Windows BitLocker, Windows PowerShell, Windows Server 2008 R2, and Windows SharePoint Services are either registered trademarks or trademarks of Microsoft Corporation in the United States and/or other countries. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.

This document is for informational purposes only. MICROSOFT MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS SUMMARY.

Was this page helpful?
(1500 characters remaining)
Thank you for your feedback
Show:
© 2014 Microsoft