4-page Case Study - Posted 7/31/2013
Rate This Evidence:
Limo Firm Virtualizes Biggest Workloads While Improving Availability, Reducing Costs
EmpireCLS Worldwide Chauffeured Services delivers celebrities and diplomats to their destinations on time and in style all over the world. Its dispatch and reservation system has to be running and performing well around the clock. While the company
had virtualized the rest of its data center, it had not virtualized this critical application—until it upgraded to the Windows Server 2012 operating system. By using Windows Server 2012, Microsoft System Center 2012 Service Pack 1, and Microsoft SQL Server
2012, EmpireCLS virtualized its dispatch and reservation system and slashed disaster recovery time from three minutes to three seconds. It shrunk its database server footprint by US$200,000 a year and IT management time by $240,000 annually. End-to-end virtualization
gives EmpireCLS far greater business agility and better performance of critical applications.
EmpireCLS Worldwide Chauffeured Services is a limousine company that offers high-end transportation services to executives, celebrities, and dignitaries in more than 700 cities around the world. But it’s also a technology company. EmpireCLS puts the
latest technology to work in its data center and in its cars to deliver outstanding customer service, reduce operating costs, and move quickly in a competitive market. In 2012, the firm started selling its proprietary dispatch and reservation system to other
ground transportation companies through a software-as-a-service model, which turned EmpireCLS into a hosting provider overnight. The company’s IT systems suddenly had to be reliable, high performance, and secure enough to support other people’s business as
well as its own.
EmpireCLS had virtualized its data centers using the Hyper-V technology in the Windows Server 2008 R2 operating system and configured its mostly Microsoft-based environment as a private cloud. This structure gave it the freedom to dynamically configure computing
resources on demand to meet ever-changing business needs.
||By using Windows Server 2012, we’ve been able to virtualize our most demanding workloads and to deliver to the business a three-second disaster recovery for these workloads at minimal cost.
| Alan Bourassa
Chief Information Officer, EmpireCLS Worldwide Chauffeured Services
The company had virtualized just about everything it could—its collaboration environment running on Microsoft SharePoint Server 2010, its messaging environment running Microsoft Exchange Server 2013 and most of its databases running Microsoft SQL Server 2008
R2. It had even virtualized desktop computers in a virtual desktop infrastructure (VDI) environment.
The one important workload that EmpireCLS had been unable to virtualize was its business-critical dispatch and reservation system, which ran on SQL Server 2008 R2 on redundant physical clusters. The application needed more memory and more virtual processors
than the 64 gigabytes (GB) per virtual machine and four virtual processors per virtual machine that Hyper-V in Windows Server 2008 R2 supported.
Alan Bourassa, Chief Information Officer for EmpireCLS Worldwide Chauffeured Services, was eager to virtualize this last holdout workload to simplify data center management and reduce costs but also to improve availability. Although EmpireCLS had its dispatch
and reservation system clusters running redundantly in three different locations to provide disaster recovery, it took two to three minutes to fail over physical servers from one location to another—an uncomfortably long time for a company that books thousands
of trips every day and has diplomats and celebrities waiting in the back seat.
In 2012, EmpireCLS upgraded its data centers to the Windows Server 2012 operating system and to Microsoft System Center 2012 Service Pack 1 (SP1) to manage its private cloud environment. At the same time, it upgraded to Microsoft SQL Server 2012 database
software. The combination of these solutions allowed EmpireCLS to finally virtualize its biggest, most demanding workload.
Scalability Improvements to Support Demanding Workloads
Windows Server 2012 Hyper-V includes significant scale and capacity improvements that were critical in the company’s decision to virtualize its biggest database workload. Windows Server 2012 supports huge host servers with up to 320 logical processors
and 4 terabytes of physical RAM and, more importantly for EmpireCLS, virtual machines with up to 64 virtual processors and 1 terabyte of memory. These improvements enabled EmpireCLS to overcome the four–virtual CPU limit in the earlier version of Hyper-V.
In addition, the company can now create more virtual machines on each host server to reduce overall costs.
With scalability no longer limited, EmpireCLS had to ensure that the solution provided the levels of performance and resiliency that it demanded for this mission-critical workload. One of the most important features in meeting this need was the new Hyper-V
Virtual Fibre Channel, which provides virtual host bus adapters within a guest (virtual machine) operating system so that an administrator can connect to fibre channel storage directly from within a virtual machine. By using this feature, EmpireCLS can not
only drive even higher levels of storage performance while maximizing its existing storage investment, but it can also build in-guest failover clusters (failover clusters that span virtual machines rather than physical servers) to support its deployment of
SQL Server 2012 AlwaysOn Availability Groups.
AlwaysOn is an enhancement of database mirroring, and an AlwaysOn Availability Group is a logical collection of databases that automatically fails over as one unit. “The combination of AlwaysOn, the Hyper-V Virtual Fibre Channel switch, and all the networking
enhancements gave us the confidence to virtualize our dispatch and reservation workload,” Bourassa says.
The networking enhancements Bourassa is referring to include Network Interface Card (NIC) Teaming, which is another Windows Server 2012 feature that helped EmpireCLS virtualize its high-end SQL Server 2012 databases. NIC Teaming provides network fault tolerance
on physical and virtual servers by using at least two network adapters by any vendor, without the need for a third-party teaming solution. Previously, Bourassa employed third-party NIC teaming solutions but could not mix and match adapters, so he forfeited
Microsoft support for a third-party-enhanced solution. With NIC Teaming now built into the operating system, EmpireCLS has a solution that is easier to deploy and manage and is fully supported by Microsoft.
||The combination of AlwaysOn, the Hyper-V Virtual Fibre Channel switch, and all the networking enhancements gave us the confidence to virtualize our dispatch and reservation workload.
| Alan Bourassa
Chief Information Officer, EmpireCLS Worldwide Chauffeured Services
In addition, with Single Root I/O Virtualization (SR-IOV), another feature included in Hyper-V, EmpireCLS can take advantage of SR-IOV-capable network cards in its Hyper-V hosts, and assign the virtual function of these devices directly to a virtual machine.
This increases network throughput, to drive higher levels of performance for key workloads, and reduces network latency, while also reducing the host CPU overhead required for processing network traffic. It’s important to note that SR-IOV with Hyper-V also
supports live migration, which means that EmpireCLS can achieve higher levels of networking performance and still have the flexibility to move key workloads around its environment as necessary, all without downtime.
EmpireCLS also uses the Hyper-V Replica feature of Windows Server 2012 to provide asynchronous replication of virtual machines between a primary and secondary site. Hyper-V Replica is agnostic of the servers, storage, and networking hardware across these
sites and provides replication flexibility and low-cost business continuity.
Continuous Availability Through Multisite Clustering
Bourassa’s staff used Hyper-V to create a virtualized test environment where it simulated the physical dispatch and reservation workload. When the team determined that the virtualized workloads performed just as well as the same workloads running in
a physical-server environment and failed over flawlessly, it migrated the physical production systems to the new virtualized environment in just two hours.
EmpireCLS runs its virtualized dispatch and reservation system on six three-node physical Hyper-V clusters—two each in three different locations. It has three SQL Server databases configured as an AlwaysOn availability group across the three sites. If the
company loses a server or an entire data center, its dispatch and reservation system (and all other applications) can be up and running in one of the other locations in two to three seconds.
EmpireCLS uses HP ProLiant BL460 Gen8 blade servers, each containing two CPUs with 8 cores and hyperthreading, for a total of 32 logical processors per blade. In addition, each blade has a minimum of 256 GB of RAM, with some hosts having 512 GB. The company
runs about 30 virtual machines on each host server, with 8 to 16 logical processors per virtual machine. EmpireCLS utilizes HP EVA P6500 storage, and utilizes the Flex-10 Virtual Connect technology of the HP C7000 chassis to present four 10-GBe NICs and Fibre
Channel storage to each blade. On each blade, EmpireCLS runs 12 instances of SQL Server 2012.
Automated Cluster Management and Performance Optimization
At the same time that it upgraded to Windows Server 2012, EmpireCLS also upgraded to Microsoft System Center 2012 SP1. The company uses the Virtual Machine Manager component of System Center 2012 SP1 to deploy and manage virtual machines, using predefined
virtual machine and service templates to provision virtual machines quickly and in an automated fashion.
With Virtual Machine Manager, EmpireCLS can also define availability sets, which enable an administrator to ensure that guest cluster nodes, such as the SQL Server 2012 Availability Groups, reside on different physical Hyper-V hosts at all times. Should
a Hyper-V host server fail, the loss does not cause the temporary loss of an entire SQL Server 2012 Availability Group.
EmpireCLS also uses the Dynamic Optimization feature in Virtual Machine Manager to optimize performance of its virtual infrastructure. Dynamic Optimization evaluates the overall performance of the cluster and, based on resource utilization and at an administrator-defined
frequency, automatically migrates virtual machines around the cluster to ensure optimal performance—without any downtime to the key workload. Additionally, Dynamic Optimization is completely aware of availability sets, to ensure that while virtual machines
are moving around the cluster, the rules that the administrator has defined won’t be broken.
EmpireCLS uses the Operations Manager component of System Center 2012 SP1 to monitor CPU performance, memory, disk utilization, and other aspects of server health and to preemptively create alerts on issues before they cause more significant problems. For
instance, if a server power supply was about to fail, Operations Manager would raise an alert. In addition, through its deep integration with Virtual Machine Manager, the two components would work together to automatically move virtual machines off that particular
host server, to a healthy server and simultaneously send an email or text alert to Bourassa’s staff.
EmpireCLS uses the Data Protection Manager component of System Center 2012 SP1 to help protect all virtual machines and SQL Server databases in the dispatch and reservation system clusters. Data Protection Manager also notes when virtual machines have been
migrated from host to host so it knows to protect them wherever they are running. The scalability and performance improvements introduced in System Center 2012 SP1 Data Protection Manager help speed up and streamline this process considerably.
The company also uses the Service Manager and Orchestrator components of System Center 2012 SP1 for help-desk ticket management and to automate complex IT workflows. For example, when the company recently deployed hundreds of Windows Phone 8 devices to chauffeurs,
the IT staff used Orchestrator to write a runbook (a series of IT workflows) to create device profiles based on the company’s Active Directory system. When a chauffeur joins the company, the chauffeur service manager can quickly and easily issue a new phone
(or disable a phone if an employee leaves the company or loses a phone) using a Service Manager portal.
|EmpireCLS runs its virtualized dispatch and reservation system on
Microsoft SQL Server 2012 on six three-node Hyper-V clusters.
EmpireCLS used Windows Server 2012 to virtualize its most demanding business application—dispatch and reservation—and in the process to improve application availability, reduce hardware and IT management costs, increase business agility, and improve
Deliver Three-Second Disaster Recovery Solution at Minimal Cost
By using Windows Server 2012, System Center 2012 SP1, and SQL Server 2012, EmpireCLS virtualized its business-critical dispatch and reservation system and, at the same time, reduced failover time for this application from several minutes to a few seconds.
“By using Windows Server 2012, we’ve been able to virtualize our most demanding workloads and to deliver to the business a three-second disaster recovery for these workloads at minimal cost,” Bourassa says. “During Hurricane Sandy in 2012, we lost one of our
New Jersey data center locations but didn’t miss a beat. Our operations continued to run from one of our other locations.” Plus, the company’s redundant backup systems are not sitting idle; they’re being used for reporting and other applications, further increasing
the value of all data center resources.
||We’ve scored major wins by using Windows Server 2012 to virtualize our most demanding workloads—cost savings, time savings, and reliability and performance improvements. We’ve proven that there’s no workload that we can’t virtualize.
| Alan Bourassa
Chief Information Officer, EmpireCLS Worldwide Chauffeured Services
EmpireCLS uses Windows Server 2012, System Center 2012 SP1, and SQL Server 2012 to gain an unparalleled level of flexibility and resiliency for moving workloads to accommodate planned maintenance or to rapidly restore workloads in the event of a disaster. ”With
all these new capabilities, we can move workloads easily and seamlessly across data centers without worrying about the physical network infrastructure and topology,” Bourassa says.
Reduce Server Costs by $200,000 a Year
By virtualizing its dispatch and reservation system, EmpireCLS has been able to shrink its database server footprint by two 42U racks, a savings of US$200,000 annually from reducing server acquisition, data center real estate, power, and cooling costs.
“We clearly have to increase server capacity as our customer base grows, but by combining HP blade servers with Windows Server 2012, we have reduced our server footprint by a factor of 12,” Bourassa says. “A 12:1 footprint consolidation is huge. Data center
efficiencies deliver a huge cost saving for environmentally conscious companies like ours. On every invoice, we tell our limo customers the carbon footprint of their trip, and we want to be continuously reducing this number.”
Reduce IT Management Costs by $240,000 Annually
Bourassa estimates that the efficiencies provided by Windows Server 2012 and System Center 2012 SP1 equate to the work of at least two full-time employees—a cost avoidance of at least $240,000 a year. “With Windows Server 2012, we don’t need as many
people or as many expensive experts,” Bourassa says. “More tasks are automated, and the tasks are simpler. We’ve reduced our IT and operating spend by 30 percent over three years, and we added 550 external vendors to our infrastructure and grew the business
by 30 percent. We couldn’t have done that without Microsoft virtualization, management, and cloud software.”
Increase Business Agility
Because the IT staff can so much more easily manage and reconfigure a private cloud environment, it has far more flexibility in configuring the IT environment on the fly. “We can add disk space on the fly to both our Virtual Fibre Channel disks and our
SAN [storage area network],” Bourassa says. “This gives us a lot of latitude in placing and moving our environment, both within a geolocation and between locations. All these operational speedups help the business move faster. IT can deliver the resources
the business needs, and change them instantly, to enable rapid business reaction and movement.”
Improve Application Performance
Performance is critical with the company’s dispatch and reservation system; if the dispatching operations team is not getting subsecond response times, they’re on the phone to the IT staff. By upgrading its dispatch environment from a Remote Desktop
Services VDI deployment of the Windows 7 operating system to the Windows 8 operating system, and upgrading the dispatch infrastructure to Windows Server 2012 and SQL Server 2012 managed by System Center 2012 SP1, EmpireCLS improved the performance of this
critical application—which was noticed by dispatch operations.
“We’ve scored major wins by using Windows Server 2012 to virtualize our most demanding workloads—cost savings, time savings, and reliability and performance improvements,” Bourassa says. “We’ve proven that there’s no workload that we can’t virtualize with
Windows Server 2012
Windows Server drives many of the world’s largest data centers, empowers small businesses around the world, and delivers value to organizations of all sizes in between. Building on this legacy, Windows Server 2012 redefines the category, delivering hundreds
of new features and enhancements that span virtualization, networking, storage, user experience, cloud computing, automation, and more. Simply put, Windows Server 2012 helps you transform your IT operations to reduce costs and deliver a whole new level of
For more information, visit:
For More Information
For more information about Microsoft products and services, call the Microsoft Sales Information Center at (800) 426-9400. In Canada, call the Microsoft Canada Information Centre at (877) 568-2495. Customers in the United States and Canada who are deaf or hard-of-hearing
can reach Microsoft text telephone (TTY/TDD) services at (800) 892-5234. Outside the 50 United States and Canada, please contact your local Microsoft subsidiary. To access information using the World Wide Web, go to:
For more information about EmpireCLS Worldwide Chauffeured Services products and services, call (800) 451-5466 or visit the website at: