Data Center Power Featured Articles
Hospital Adds SDN and Network Virtualization with Help from Nuage Networks
Last week the University of Pittsburgh Medical Center (UPMC) selected Nuage Networks software defined networking (SDN) solutions for deployment in the health system’s backup network. This deployment is part of the hospital’s multi-phase transition to SDN and network virtualization.
Nuage Networks is an April 2013 spin-out from Alcatel-Lucent (News - Alert), with a focus on SDN solutions. Last year it announced its virtualized services platform software, which included a directory service for policy and analytics, an OpenFlow-capable network controller and a virtual and physical network forwarding plane.
The Medical Center has recently completed phase one of this multi-part transition, which replaced a legacy Layer 2 network. This was being used as a dedicated backup system designed to carry traffic between servers and a tape library system.
Currently, the Nuage Networks solution is being used on the hospital’s backup network. If all goes as proposed and a staging and verification period completes without any problems, UPMC will begin a multi-year transition to Nuage Networks solutions for the rest of its datacenter network infrastructure. This will support the increased demand from employees, patients, hospitals, and healthcare insurers, according to Bill Hanna, VP of technical services for the University of Pittsburgh Medical Center.
Phase one was described as Hanna’s team ripping out the legacy Layer 2 backup network to make way for Nuage hardware. This consisted of four 7850 Virtualized Services Aggregators in the spine and eight Virtualized Services Gateways as leaf devices. This hardware is meant to provide the physical substrate over which Nuage's network virtualization overlay product, the Virtualized Services Platform, operates. This also forms a leaf-and-spine network.
The next phase sees Hanna’s team ripping out most of its legacy Cisco (News - Alert) and Alcatel-Lucent switches. This will allow production traffic to run over the Nuage hardware. Cisco Nexus 7000 switches will remain in place to route traffic in and out of the data center.
There is still a lot that needs to be done. Training and retraining will be required for the engineers and operators. The reason according to Hanna is "because they're command-line jockeys." Upon completion of the training, he expects the provisioning time for network services in his data center to drop dramatically.
Hanna said, "We want to prove it out. When the time is right and everything works well, we'll start transitioning more traffic to [the Nuage] network for production. We'll take that Nuage network and hook it into the rest of our data center infrastructure -- mostly Cisco Nexus. We'll tie it into the Nexus 7000s, [which] will make routing decisions."
The following is why Hanna expects to see a dramatic drop in provisioning time, he said, "From a provisioning standpoint, we have the data communications people, then we have the people who do hosts and the people who manage the address space and the DNS, and then we have people who manage the cabling. Our hope is that … it will end up being an hour moving forward."
Apart from the “ripping out” of equipment, Hanna seems very excited about this move. He believes that his networking team will enjoy the flexibility of a virtual network. Nuage Networks will give his team the ability to regain control over the edge of the data center network.
Network engineers should enjoy the fact that they will have more insight, as well as control over the virtualized edge of the network. This is accomplished through Nuage’s VSR endpoints on each VMware host.
Following up on this, Hanna said, "Once you hand off a port to a VMware host [in a legacy network], those network guys have very little insight into what's going on inside that box. Nuage allows multiple people to do provisioning in the network. Today we move VMs all over. That's one of the problems we're solving. It gives you flexibility. Look at VMware [vSphere]. It is very flexible. You can quickly move things around. The traditional network is not that flexible. We're solving that problem."
During the upcoming SDN Precon on Jan. 28 in Miami, Fla., attendees will hear feature presentations and gain educational opportunities on everything SDN.
Edited by Rory J. Thompson
Data Center Power Resources
Featured White Papers
As the need to balance current and future IT requirements against resource consumption becomes more urgent, the data center industry increasingly views capacity planning as a way of achieving a critical component to planning a new build or retrofit. Data center capacity planning can be a complex undertaking with far-reaching strategic and operational implications. DCD Intelligence has therefore compiled this White Paper in order to share some industry insights and lessons on the practical steps that are needed to develop a successful power and capacity planning strategy.[Read More]
Server Technology had the recent opportunity, along with other partner companies, to participate in discussions across the globe with data center IT and facility managers as part of a road show seminar: Data Center Energy and Operational Efficiency.[Read More]
The demand for more power in the computer cabinet has led many data centers to upgrade to three phase power distribution. Proper three phase power distribution has traditionally meant dividing up power up into multiple branches within the rack PDU (Power Distribution Unit). In this paper we will explore the advantages of a new, less common approach to PDU design by means of alternating each phase on a per-receptacle basis instead of a per branch basis.[Read More]
Increasing powering and cooling demands within the data center have been the topics of choice for Data Center (DC) and Facility Managers for several years now. Increased power demands are a result of the need for more compute power and higher density devices have resulted. These high density installations include stacks and stacks of servers and the trend of implementing blade servers within these server "farms." Cooling problems are a direct result of the increased power demands based on the simple fact that more power increases the demand for cooling.[Read More]