I just got back from Carrier Ethernet World Congress 2011 in Amsterdam and it provided a great opportunity to meet many of our service provider customers from around the world. It also proved to be an ideal forum to share experiences as well as visions for the communications industry between different vendors and network operators. For the Cisco team, one of the highlights of the event was winning the “IIR Carrier Ethernet Vendor Award EMEA” in the “Best Carrier Ethernet Aggregation Product” category with the Cisco ASR 9000 System. Now deployed with over 750 customers worldwide, this was another great endorsement of our Carrier Ethernet strategy and the cost benefits associated with Cisco Network Virtualization (nV) technology.
MPLS 2011 Conference in Washington D.C., is two weeks away; and we can anticipate a dynamic venue that commences with a Cloud Architecture tutorial to be presented by my Cisco colleagues Azhar Sayeed and Kapil Bakshi on Sunday October 16.
Monday, October 17
Cisco’s Clarence FilsFils will present Data Center Interconnect: Multi-Homing, Resiliency and Mobility; to be followed in the afternoon by my presentation on SP Cloud Computing: Services, Architecture and Standards Considerations and Cisco’s John Evan’s presentation on Application and Service Aware IP/MPLS Networking.
To end the day, I will chair a panel on the Role of MPLS and Ethernet Network Technologies as Enablers to Cloud Computing with the participation of outstanding individuals in the industry such as:
- SERGEI GERASIMTCHOUK, Facebook
- SCOTT WHYTE, Google
- VIJAY GILL, Microsoft
- GUENTER HONISCH, Deutsche Telekom
- ASHWIN GUMASTE, IIT
- NING SO, Verizon Read More »
Cisco’s foundation for delivering the service provider Cloud is our Unified Service Delivery (USD) solution, featuring tightly integrated, data center and IP NGN technologies to deliver a virtualized end-to-end infrastructure for cloud services. We thought it would be useful to share some new capabilities that Cisco has added recently to the solution:
MPLS in the Data Center: To streamline the end to end operation across the data center and IP NGN, Cisco announced, last week, that Nexus 7000 supports Multi Protocol Label Switching (MPLS) for Layer 3 virtual private networks (VPNs). This allows MPLS to be deployed at the data center core/aggregation layer rather than terminating at the data center edge. This capability enables Service Providers to greatly simplify L3 segmentation, especially for multi-tenant cloud offerings, depending on their scale and service needs.
Another key to the delivery of a data center built for Cloud requirements of scale, virtualization and multi-tenancy has been the use of a Unified Fabric. Unified Fabric provides the flexibility of high performance, highly available networks to support the needs of both LAN and SAN on a consolidated fabric. Cisco’s Unified Fabric announcements last week bring new capabilities which extend our already robust offerings to further build out a Service Provider Cloud foundation.
There has been a lot of buzz recently about a second OAM (Operations, Administration, and Maintenance) solution for MPLS-TP that will cause interoperability problems between MPLS-TP and MPLS. It is accurate that there is an alternative OAM based on ITU-T Y.1731 (Ethernet OAM) proposed by a number of vendors and countries and indeed, it will cause interoperability issues. As a strong believer in standards, I certainly hope that a second approach does not occur because vendors and customers do not need the additional cost burden that a lack of interoperability causes. The fact is that only the draft recommendation for MPLS-TP OAM based on Y.1731 has begun the first step in a very long approval ITU process -- but nothing more – and in my estimates will take well over a year and could easily take up to two years to standardize. IETF MPLS OAM is widely deployed in MPLS networks today and will simply be extended as MPLS-TP is deployed as a next generation transport solution. In fact, recent interoperability testing of MPLS-TP took place at the MPLS World Congress earlier this month in Paris.
I believe that after careful consideration most operators will see the benefit of having a single end-to-end methodology to operate and manage converged packet optical transport networks, which MPLS-TP using MPLS OAM provides. Operators who select another method that is perceived to meet their short term needs now my ultimately learn that it fails to provide everything they had expected, and that having multiple OAM methods (one for Ethernet and another for MPLS) is not cost effective. It will be interesting to see what happens moving forward. At the very least, operators should make an informed decision on which approach is right for them.
SONET and SDH, while well established and highly reliable technologies, were optimized for an era of TDM voice communications - before cloud computing, VoIP, cell phones, mobile tablets, Video CDNs and even the Internet increased network complexity and radically changed traffic patterns. This infrastructure has delivered reliable transport because it is easy to provision, troubleshoot, and provides a high level of resiliency. However, with the increase in IP traffic and changing traffic patterns, a new solution is required: one that provides the ‘trust’ of SONET/SDH with the ‘efficiency’ and ‘agility’ of packet.
Today Cisco is announcing our latest innovation, the Cisco Carrier Packet Transport (CPT) System. Our goal is to enable the transport network in a way that combines the reliability and simplicity of point-and-click provisioning of SONET/SDH along with the efficiency and flexibility of IP/MPLS.
Service providers generate a lot of revenue from connection-oriented services like leased lines. However the growth of these services from a revenue perspective is slow. New services based on cloud, mobile, and video are IP based and have huge growth potential. Service providers need to address this growing traffic and need to do so profitably. They need to find ways to lower the cost of transport and simultaneously tap into new applications that increase the average revenue per user. The Cisco Carrier Packet Transport System helps them do exactly that!
Of course, new technology can be overwhelming. So, it’s important to deliver all the benefits of packet technology without sacrificing the trust of transport. With the Cisco CPT System, service providers can build a packet transport infrastructure with the same reliability and familiar operational models of SONET/SDH. Standards-based MPLS-TP allows for robust packet connection-oriented control. Cisco’s Premier Integrated Management Experience (PRIME) offers service providers an A-Z point-and-click network management system. Both wavelengths and MPLS-TP label switch paths are provisioned in an easy “point-and-click” fashion autonomously or from a single integrated domain. This integrated solution provides opex savings and eliminates the need for overhaul and extensive employee re-training.