Welcome to the Cisco Networking Professionals Ask the Expert conversation. This is an opportunity to learn how the Unified Service Delivery solution transforms the way service providers deliver services with Cisco experts Curt Equi, Wayne Cullen and Brian Korn. Curt Equi is a senior manager in Cisco's Service Provider marketing organization, with responsibility for the Unified Service Delivery campaign. He has more than 15 years of experience in the enterprise and service-provider markets across a wide variety of technology disciplines, most recently focusing on data center solutions. Curt holds a bachelor of science in applied mathematics and computer science from Carnegie Mellon University. Wayne Cullen is responsible for solutions marketing to service providers worldwide. His current focus is on the Cisco Unified Service Delivery campaign, which encompasses infrastructure solutions for service provider data centers, their emerging cloud computing services, and other forward-looking approaches to service delivery. He has more than 25 years experience in the telecom/datacom industry spanning development engineering, manufacturing, marketing, operations, and business consulting roles. Wayne holds a bachelor of engineering degree from the University of Queensland, Australia. Brian Korn is a marketing manager at Cisco Systems responsible for Unified Service Delivery and Cloud for Service Providers. Brian holds a bachelor of science degree from Lehigh University and an MBA from the Kellogg graduate school of management at Northwestern University.
Remember to use the rating system to let Wayne, Curt and Brian know if you have received an adequate response.
Wayne, Curt and Brian might not be able to answer each question due to the volume expected during this event. Our moderators will post many of the unanswered questions in other discussion forums shortly after the event. This event lasts through August 28, 2009. Visit this forum often to view responses to your questions and the questions of other community members.
What's the difference between the Cisco Unified Service Delivery and the Cisco Unified Computing System? From what I can see one is for service providers and other for enterprise -- can they be used beyond these perimeters?
The Cisco Unified Computing System (UCS) is Cisco's multi-chassis blade computing system based on Intel Xeon 5500 processors and lossless 10Gb unified fabric (FCoE and 10Gb Ethernet functions). UCS incorporates simplified management and service profiles to facilitate dynamic system provisioning.
more on UCS is available at www.cisco.com/go/usd.
Unified Service Delivery (USD) unites the datacenter and IP-NGN for service providers to create a common infrastructure from which services can be deployed in a secure virtualized fashion. The Unified Computing System is leveraged by USD to deliver physical and virtual servers required by a service.
More on USD is available at www.cisco.com/go/usd
The data center and the network are two very different environments. What exactly does Cisco mean by âunitingâ them?
Hi, Jean. You're right... the datacenter and the network are *very* different environments. And we're not proposing to unite them in terms of making them one and the same.
Rather, what we're talking about is unifying the way providers deliver services across both of these very different environments.
What do we mean by that?
Well, have you ever noticed that there is sort of an "invisible line" between the network and the data center, and that it's difficult to make decisions about solutions or technology that span that line? In other words, decisions are often made from the perspective of network operations, or based on data center requirements, but rarely take the needs both environments into account equally.
This isn't a criticism of the decision-making process, it's just an observation of the reality of many service delivery operations. Perhaps the network has historically been the core asset of the provider (as with a traditional carrier) and the subscriber-facing datacenter is seen as secondary to that. Or maybe the provider delivers primarily datacenter services and the situation is reversed. Or maybe the provider operates a VAN provided by a carrier partner.
And we technology suppliers haven't helped the situation. There are "data center" solutions and "network" solutions, and we market them as such. In fact, vendors in our industry are often characterized by what environment we play in most strongly... we are either a data center vendor or a network vendor. We have often approached our customers with solutions for one or the other of these environments, but rarely both. That contributes to the "invisible line" problem, and has to change.
With Unified Service Delivery, Cisco is making that very change. We are delivering solutions that take the requirements of both the IPNGN *and* the SP's subscriber-facing datacenter into account. Our technology in the datacenter will work in concert with our technology in the network to create a "1+1=3" effect where those two assets of the provider combine to deliver more value together than either would alone.
By way of example, let's take a look at Cisco is doing with virtualization on our core routers and core datacenter switches. For some time now we have had Secure Domain Routing (SDR) on the CRS-1 carrier-grade core routing platform. This gives network operations staff the ability to provide complete logical and physical separation of resources on the routing platform and throughout the network. Problem was, that this separation stopped once it exited the WAN and entered the datacenter. Suddenly, providers either needed to collapse services onto a single switching platform in the datacenter and lose separation, or provide multiple separate hardware platforms - wasting space, power, and money - to achieve true separation.
Our answer to this is Virtual Device Contexts (VDCs) on the Nexus 7000-series of 10G core datacenter switches, in conjunction with SDR. VDCs can provide the same type of physical and logical separation of resources on the datacenter switches as we have provided with SDR on the CRS-1 routers. In this way, service separation can be efficiently achieved throughout *both* the network *and* the datacenter.
This is just one example of how Unified Service Delivery will help providers deliver services more effectively and give them the flexibility they need to create new and more differentiated services.
As I stated earlier, we're not uniting the IPNGN and the datacenter in the sense of making them one in the same. Rather, we're helping providers deliver services across both environments in more of a unified fashion so that the individual strengths of both of these assets combine to deliver more than either would alone.
Hope this clarifies things for you!
Hi, On ANM v2.1 is there any option to find connection established of perticular IP as we can see in CLI ? Please suggest if any setting in GUI for that?
If you mean the statistic as to number of connections for a VIP or a Real server, and not assuming which statistic you mean, then in ANM 2.1 you may be finding this on the Operations pages. On those pages you can use the filter/sort functions (upper right on screen) to find for a particular matching (set of) IP address. Similar and additional data is available in the monitoring tab under the load balancing selection.
In ANM 3.0, currently in development and target for delivery in Q1CY10, a significant enhancement to the monitoring pages.
Feel free to also ask such questions at firstname.lastname@example.org.
i am talking about how to see all static connection made to VIP, its state & to find out connection for particular IP address has been made in ANM
Within ANM you can only see the aggregate connectoins per VIP, not broken down to track a particular IP address of a client connecting to that VIP.
You can track IP connections on a VIP (and denials) via the Syslog functions on ACE itself.
Our big challenge in our data centers is power density, high costs of power and cooling. How does the Cisco USD do in regard to power consumption, cooling requirements, and power density in the data center?
Your challenge is shared by most, if not all, data center operators worldwide. Much of the problem lies with poor efficiencies in existing data centers - too much I/O overhead, too little utilization of servers, storage, networks. Cisco has approached it in a holistic manner ( www.cisco.com/go/usd ) not just one aspect which would limit a solution. Thus we believe we can offer significant improvements across many issues in data centers - better utilization, lower administration costs, and of course the core to your problem much improved power and cooling efficiency.
Our Unified Fabric offer, http://www.cisco.com/en/US/netsol/ns945/index.html can eliminate the three networks environment of typical data centers thereby reducing I/O overhead, which has a direct effect on lowering power consumption 30% or more, improving rack space utilization, improving power density per compute resource, eliminating much cabling and improving airflow for cooling and more. By unifying the compute resources with the network, http://www.cisco.com/en/US/netsol/ns944/index.html we further improve power and cooling efficencies by eliminating overhead switch ports and management blades and have designed a system which has greater cooling attributes than alternatives. To put some real-world numbers around this using our own data center experience with Unified Service Delivery approach - in our data centers that have implemented this approach, we now support 4 times as many VMs per kW than before, reduced rack space to almost a half of before, power savings due to I/O reduction has resulted in 30% more power available for servers for the applications. Analyst firm Synergy also did an assessment of the Cisco Unified Service Delivery and which can be found at http://www.cisco.com/go/usd . They concluded that the Cisco USD improved power efficiency by a factor of 2.