Cisco’s Cloud Orchestration Platforms Showing Increased Versatility and Strength

As providers more clearly understand the fundamentals of cloud computing (in areas such as activating virtual machines and allocating storage pools to support them) they are turning their attention to and focusing on broader capabilities in cloud service creation and management. Expanding capabilities in these areas is crucial for clouds to become the scalable engines of service delivery the industry envisions. Developers’ energies are also focused on streamlining service creation, flexibly supporting multiple cloud combinations, and integrating cloud service creation with broader IT service management to make leveraging the cloud easier to achieve.

Cisco has been steadily assembling an array of cloud service management capabilities over the past year, and this week at Cisco Live! a number of them were on display. They demonstrate that Cisco’s efforts are bearing important fruit. Three areas showing substantive progress toward important service management goals stand out: automatic integration of multiple clouds into single service offerings, automatic installation of application profiles into target virtual infrastructures, and architectural flexibility from integration of OpenStack and other platform services.

Cisco’s Intelligent Automation for Clouds (CIAC) portfolio allows managers to consolidate many clouds into one by incorporating support for public cloud services, such as Amazon’s, Verizon’s, and HP’s, into multicloud service provisioning that can also include resources from an organization’s own infrastructure, such as ESX, KVM or Hyper-V VMs. The process enables service managers to automatically blend functionality from any of the supported offerings into a whole service offering for users. For most organizations this capability will be table stakes over time for realization of cloud deployment goals.

A second area of equal importance is streamlining the relationship between application development and operations in a target environment. There is a growing premium placed on shortening the time between creating a new application function and being able to put it in users’ hands. Vendors are developing tools that support this process and help cloud service managers achieve this creation and delivery goal. The challenge is how to successfully combine development and test and ultimately operations in well-defined execution environments and tightly coupled workflows to reduce the time between application innovation and application use.

Cisco is growing its capabilities in this area through work with its partners Puppet Labs and OpsCode by integrating these partners’ devops platforms with its own service automation platform. In doing so Cisco’s platform leverages the application profiles developed in both Puppet’s and OpsCode Chef’s environments and incorporates them into the cloud service management environment it is enabling. This again shows new muscle in creating cloud management capabilities well-suited to the agility and pace of innovation required in the virtual enterprise.

The third indicator of growing versatility and strength in the CIAC portfolio is expanded support for OpenStack software distributions. This is important because as suppliers bring capabilities to market in areas where they have unique strengths to offer (such as harvesting network analytics in support of service optimization in Cisco’s case) there may be other cases where a seamless integration of another supplier’s capabilities via a platform like OpenStack will bring a richer environment into play for the customer. For example, an OSS application may uniquely leverage OpenStack’s emerging Ceilometer metering functionality to create distinct service bundles in a multicloud environment, and facilitating integration of this application with other service management capabilities may be enabled by leveraging the modules and API suites of the CIAC environment. Support today in CIAC for OpenStack’s Nova (compute) and Horizon’s (dashboard) APIs is the first step toward realizing this objective and show strong insight about what the operating requirements of many cloud computing customers will be. Eventually, the opportunity to provide a truly versatile service creation platform for enabling both innovation and efficiency may very well emerge.

In each of these areas Cisco has shown it is making tangible progress toward its goals of excelling in the world of multiple clouds and enabling the open and elastic services required for success in the new virtual enterprise. If you were using a compass to gauge the utility of these developments for virtual IT, the indicator would be reading direction correct.

For more information on ACG's cloud services, contact sales@acgresearch.net.

Paul Parker-Johnson
pparkerjohnson@acgresearch.net
www.acgresearch.net

Vendor Commitment to SDN in the Optical Environment

Not wanting to create the perception that their solutions are behind the times and anything less than state of the art, the telecommunications equipment vendors are always fast to embrace the next thing. However, the reality is that adopting new technology/methodology into diverse multivendor elements that comprise a service provider’s (SP) network is a significant undertaking. This has caused some of the larger content service providers such as Google and Facebook to undertake the building and deployment of their own private infrastructures to support their businesses. Because users’ demand for content delivery is outpacing the cost per bit to deliver the data, traditional SPs are getting squeezed. Simultaneously, the time to turn up a service offering has become a real and competitive advantage. This has forced SPs to look toward new approaches such as software-defined networking (SDN) to reduce service turn-up times and better leverage the infrastructure to support content delivery and achieve CapEx and OpEx benefits. But for SDN to deliver on these benefits it must work in a multivendor environment and end to end across all services supporting elements in the network.

Currently, this level of SDN deployment has only been achieved in a couple of private networks with proprietary implementations undertaken by the providers. Although this validates the need and benefits of SDN it by no means makes it mainstream. The question remains, how long will it be before SPs can implement SDN in a key portion of their networks such as optical transport? Examining some recently published vendor activity indicates just how ready for prime time this technology really is.

April 11, 2013, Ciena to Showcase Service Provider SDN at Open Networking Summit. Ciena demonstrated two service scenarios that leveraged the automation and central intelligence of its OPⁿ network architecture to show automated provisioning, virtualization and bandwidth on demand.

June 12, 2013, Cyan to Demonstrate the First SDN Application Spanning Enterprise, WAN and Data Center Environments at Interop Tokyo. This demo tested several use cases and showed interoperability between vendors. Cyan’s goal was to demonstrate the virtualization of the data center and network resources. It included members of the recently formed Blue Orbit Ecosystem

June 18, 2013, Coriant announces Intelligent Optical Control (IOC), industries first solution advancing SDN for optical networks. Coriant claims its solution, which is the first, allows for the optimization of the optical portion of the network. Its solution reduces CapEx by as much as 50 percent.

June 24, 2013, Infinera Demonstrates Transport SDN and Packet Technology on DTN-X Platform at Nissho Labs. This demo featured the DTN-X working with an external SDN controller and different network applications. It provisioned bandwidth on demand using OpenFlow and included VLAN switching and MPLS pseudo-wire transport over a 500G super-channel.

Although these press releases are show demos or lab trials they are key indicators of market direction and vendor uptake — to get to the demonstration stage the equipment vendors have invested in development resources to achieve this level of interoperability. The testing also helps harden the solution as nuances are identified and addressed by the vendors; it is a gauge of the technology maturity level of the solution. Multivendor environments must operate end-to-end to receive the full value and promise of an SDN networking environment. At this stage, at least for the optical transport portion of the network the equipment, vendors seem very committed to SDN and are poised to begin delivery of SDN ready systems in earnest as soon as 2014.  

For more information about ACG Research’s optical services, contact sales@acgresearch.net.

           Jeff Ogle

jogle@acgresearch.net   

     www.acgresearch

Takeaways from the Cable Show 2013

It was apropos that the 2013 Cable Show was held in Washington D.C. considering the big role that regulators play in the industry. Not only did attendance seem lighter than last year, but opening sessions on Day 1 also were disjointed with speeches that were self-congratulatory, lacked substance, and made claims that were hard to believe, such as “the industry is an innovator; the US is really a leader in average Internet speeds.” And “all content will be available on all screens regardless of what MSOs and programmers think, and today’s disruption is really about business models.”

In Day 2 Michael Powell’s (president and CEO of NCTA) interview with Mignon Clyburn (current acting chair of the FCC) was interesting from a human perspective but lacked any real content. Brian Roberts demonstrated the Xi3 box and the new X2 user interface for Xfinity, which include voice commands and seem to be as good if not better than Siri. It appears to be powered by Veveo. I was impressed because it seems that Comcast finally got religion about user experience; it will be interesting to see how fast they can roll it out and how well it will actually works in the home.  

Other themes

M2M/IoT: CPE vendors and service providers were demonstrating home security, monitoring and control capabilities—a huge growth opportunity and a natural extension of their existing markets.  

Cloud: Cisco demonstrated cloud DVR (http://www.acgresearch.net/knowledge-insights/videos/cloud-dvr-perspective.aspx). ActiveVideo demonstrated cloud-powered guides. ActiveVideo showed that mission-critical functions can be moved to the cloud. The company also announced wins with Charter and Cablevision.

CCAP: Cisco, Arris/Moto and Casa demonstrated their capabilities, and we are starting to see commercial shipments, which should continue throughout the year. 

Gateways are starting to gain traction as well. Cisco was giving a sneak peak at its G8 hybrid IP gateway. Arris and Technicolor also were showcasing their products.

Azuki Systems, an innovator in delivering multiscreen over legacy networks, unveiled an approach for using its technology to reduce the bandwidth requirements for first screen viewing to enable QAM reclamation.

For more information about ACG Research's VideoInfrastructure service, contact sales@acgresearch.net.

David Dines

ddines@acgresearch.net
www.acgresearch.net

2013: The Year of Over the Top

Video continues to be the major driver of bandwidth demand on fixed and mobile networks. According to a report by Sandvine, it consumed 62% of peak time fixed Internet bandwidth and 43.5% of mobile bandwidth in North America. In other regions, viewership is lower but is still significant and growing.

H.265/high-efficiency video coding (HEVC) capabilities are just starting to enter the market. We anticipate that adoption will be faster in mobile because of the rapid turnover of devices and the capacity limits of the RAN. In wireline, it will take longer to see mass adoption, as operators balance the efficiency and service gains against the costs of replacing significant amounts of CPE as well as the encoding infrastructure. The other argument for HEVC is the ability to offer new, ultra-high definition (UHD or 4K) service. We believe the jury is still out regarding consumer interest in UHD because of the extra cost of the sets, lack of available programming and perceived improvement in viewing experience.

TV everywhere or multiscreen viewing is now accepted by the SPs as table stakes in providing video. The issue that SPs are starting to address is not so much supporting tablets and connected TVs but identifying what content is available and improving the users’ experiences in search and discovery. 

Competition from over-the-top (OTT) players Netflix, Hulu and YouTube continue to post impressive traffic gains. Netflix continues its quest to acquire original and premium content. It is adding premium children’s programming, acquired rights to “Arrested Development” and just inked a deal with DreamWorks Animation for approximately 300 hours of new television episodes. This deal will be closely watched by both OTT incumbent pay TV operators and other OTT providers such as Amazon, Hulu and YouTube. Although the incumbents have not yet felt a significant impact of OTT in cord cutting or cord shaving, this new deal will likely turn up the heat. 

ACG has been consistently saying that OTT is a disruptive technology/business model and that it would eventually resolve the issues that slowed adoption early in the life cycle (low quality of experience and lack of compelling content). It appears that 2013 is going to be the year that OTT comes into its own. 

For more information about ACG Research's Video Infrastructure service, contact sales@acgresearch.net.

David Dines
ddines@acgresearch.net
www.acgresearch.net

Declines in Most Segments of Video Infrastructure Market

The Worldwide Service Provider Video Infrastructure total market was down 7.2% Q-Q and 10.3% Y-Y. The declines were in nearly all segments with only core routing and CMTS with sequential increases this quarter.

Cable and IPTV STBs declined because of slowing subscriber growth and ASP declines. CMTS gained 6.4% quarter-Q but declined 26.2% Y-Y. The decline was caused by volume shipments of higher density line cards and software upgrades to existing line cards, which dramatically lowered average selling prices. Packet infrastructure posted an annual decline in core routing and was nearly flat in edge routing.

Not surprisingly, Cisco is the dominant player in the space because of its broad-based coverage of the market and is the only vendor with strength in routing, switching and CMTS. Cisco has also been leading the IPTV STB market for several years and took the lead in cable STBs from Motorola.

1Q/13 Worldwide Service Provider Video Infrastructure
Company	Rank	Market Share
Cisco	1	48.7%
Motorola	2	13.7%
Alcatel-Lucent	3	7.7%
Pace	4	6.8%
Juniper	5	4.7%

1Q/13 Worldwide IPTV STBs
Company	Rank	Market Share
Cisco	1	42.0%
Motorola	2	14.6%
Skyworth	3	7.2%
ADB	4	6.7%
Pace	5	6.6%

1Q/13 Worldwide Cable STBs
Company	Rank	Market Share
Cisco	1	31.3%
Motorola	2	31.2%
Pace	3	12.1%
Humax	4	6.0%
Skyworth	5	4.7%

For more information about ACG Research's VideoInfrastructure service, contact sales@acgresearch.net.

David Dines
ddines@acgresearch.net
www.acgresearch.net

Ciena, Infinera and NSN Gain Market Position in 1Q WW Optical Long Haul DWDM Market Segment

In Q1, 2013 the worldwide revenue for the Long Haul DWDM market dropped from its average quarterly run rate of approximately $1 billion to $782M, a precipitous drop of 26.3% over the previous quarter yet 22.1% higher on a year-to-year basis. Y-Y this is the highest performing segment within the optical market segmentation. The top five players accounted for 83.1% of the available Long Haul DWDM segment market; however, one major difference in the 1Q optical market performance is the ranking of the more pure play optical vendors over the more traditional multitechnology providers. In 4Q only Huawei and Alcatel-Lucent made the top five. In 1Q Ciena, Infinera and NSN (Coriant) all advanced at least one rank; Infinera gained two ranks. 

1Q/13 Worldwide Optical Long Haul DWDM Market
Company	Rank	Market Share
Huawei	1	29.7%
Ciena	2	16.7%
Infinera	3	13.1%
Alcatel-Lucent	4	12.5%
NSN (Coriant)	5	11.1%

Market Drivers and Forecast

In addition to its traditional role in Metro, Long Haul and mobile backhaul networks, the optical market, in general, is also gaining traction within cloud and data center networks. Optical vendors have integrated their classic SONET/SDH, DWDM and packet optical products to offer single hardware architecture with scalable platforms targeted for different network deployment points that effectively maximize the solution for their installed bases. The more pure play optical vendors are also increasingly enhancing and adding to their transport capabilities with features such as fast rerouting and adding level 2.5 protocol support such as MPLS to climb the stack. These features help with network resiliency and provide more functionally and service deployment options to a carrier.

100G WDM is becoming the de-facto standard for Long Haul transport, driven by the need to support multiple 10G and 40G subscriber connections. The optical vendors are already working on ways to combine multiple 100G long haul connections to form super channels that will enable transport rates into the half terabit and ultimately full terabit line rates. This should better position the optical transport solutions against router technology, which is more expensive for higher speed interfaces.

With the demand for bandwidth continuing to increase and optical vendors adding value to their platforms to become better transport enabled, AGC predicts the Long Haul DWDM optical market segment will continue to grow 20–25% in 2013. 

For more information about ACG Research’s optical services, contact sales@acgresearch.net.

         Jeff Ogle
jogle@acgresearch.net   
    www.acgresearch

Composers Increasingly Important in Cloud Orchestration Frameworks

The value of cloud computing is best realized when services function transparently and automatically at any scale. Although this nirvana goal is clearly in sight for both service developers and providers, platform offerings are moving toward it in progressive stages of innovation, arriving at an inflection point now in which new capabilities housed in service composer applications are moving us closer to the goal.  

Early developments toward cloud computing in the last four to five years have provided good capabilities for creating virtual infrastructure resources, mostly on an a la carte basis. For example, systems have been developed to create virtual machines, storage pools, and virtual networking services typically within a single cloud architectural context such as VMware’s ESX and vCloud, Citrix’s XenServer and CloudPlatform, and Red Hat’s KVM and Enterprise Virtualization. Although there are resources to help stitch pieces together and some stages of orchestration across resource types to simplify provisioning, these have been largely focused on creating the pools and less on constructing whole services.

These foundations have been crucial in setting the baselines on which many XaaS capabilities available today have been realized. However, on their own they do not accomplish the goal of automating design and composition of whole services from multiple infrastructure categories or even multiple clouds. There has been significant progress made toward this goal in the past year, and we are now entering the composer applications phase, which provides new capabilities on top of resource provisioning to accelerate service design and delivery.

Composers are about creating offerings at the level of a consumable service, for example, burst computing, elastic archiving or application development integrating the services of multiple clouds. Composers assemble a service from multiple contributing sources and deploy it to a target user community. These sources can be virtual pools from within an individual cloud, combinations assembled from the resources of multiple clouds (private or public), and software to support important functions such as metering, monitoring, and SLA support.

These expanded capabilities can be seen in many deployment scenarios. Here are just two:

1.  A corporate IT group creates an on-demand XaaS offering (on-demand compute or elastic storage backup) in a private cloud. The service is offered consistently across diverse user groups (multiple business units or functional areas). The common offering is realized from the underlying resources of two or more virtual infrastructure platforms, for example VMware vCloud and Red Hat Enterprise Virtualization. The composer simplifies and streamlines the job of integrating the facilities of the two virtualization environments into a single XaaS capability for the company.  
                                                                                                   
2. An elastic storage archiving service is developed that blends enterprise on-premise and public cloud resources transparently. Designing the service is simplified by a composer app that automatically combines the private and public resources into a complete service offering more rapidly than if the combination had to be hand tooled. The service provides cost and scaling benefits to the organization, and the entire deliverable is produced more efficiently.

Examples of systems providing composer capabilities include IBM’s SmartCloud Orchestrator, HP’s Cloud Service Automation suite, Dell’s Enstratius Cloud Management platform, and Flexiant’s Cloud Orchestrator system. This year expect the Open Stack Heat orchestration project (currently under development) to yield an open-source platform for multicloud service composition (in conjunction with other functions it will provide).

The upside of this collection of developments for providers of cloud computing services is greater efficiency and speed, and increased agility in deployment of services. Leveraging the automation of a composer platform may improve speed and efficiency on the order of 50 percent or more, depending on the details of the environment. The opportunity to achieve improvements of this size in design and delivery efficiencies elevates the importance of composers in the toolkit planning providers of cloud services need to do for creation of their packages.

With this new level of service design and deployment functionality emerging at the upper layers of operators’ service management environments we can anticipate rapid production of a richer mix of applications and services. And because of their ability to accelerate time to value as well as the speed of innovation, we can expect this new powerful class of composer applications to occupy a place of prominence in operators’ plans for their cloud orchestration systems. 

For more on cloud, read Hybrid Cloud Networks Face Challenges, but SDN May Be the Answer.

Paul Parker-Johnson
pparkerjohnson@acgresearch.net
www.acgresearch.net

Cisco Captures Pole Position in 1Q 2013 WW Mobile IP Infrastructure Market

1Q 2013 Worldwide Mobile IP Infrastructure market grew to $1.2B as Mobile SPs continue investments in Mobile IP Backhaul and Packet Core networks, including significant CapEx spend shift to LTE networks. This phenomenon has been prevalent in NAM for 12 months and is now taking shape in APAC. EMEA Mobile SPs are beginning to confidently plan LTE investments as the regulatory environment becomes more favorable.

Mobile IP Infrastructure Worldwide Market Shares Q1/13
Company	Rank	Market Share
Cisco	1	42.0%
Ericsson	2	15.5%
ALU	3	13.0%
NSN	4	7.1%
Huawei	5	5.8%

Cisco has executed well with significant market share gains in 1Q 2013, achieving pole position to lead all three segments: Mobile IP Backbone, Mobile IP Backhaul, and Packet Core at total 42 percent WW Mobile IP Infrastructure market share. Cisco has also out-executed Ericsson’s stated “seasonal weakness,” taking the #1 position in the coveted EPC segment. Although Cisco has focused its Packet Core offerings with a high price/performance portfolio strategy, its sales execution within the US, Canada, and EMEA markets is yielding strongholds in key LTE markets with wins in Vimpelcom (Russia), SFR (France), select T-Mobile properties in Europe, as well as with mobile network expansions in Bell Canada, du, Bharti Airtel, Tata, and KDDI. Cisco’s core strengths in Packet Core and its deep LTE core network experience are a result of its large/incumbent position in AT&T and Verizon Wireless where it has acquired complex deployment experience in 3G migration, LTE capacity planning, subscriber policy management, and multimedia/video revenue creation models.

Ericsson, despite its 1Q 2013 weakness, maintains its “Game of Thrones” empire on LTE networks. Massive global LTE deployments and a high rate of trial contract conversions/expansions, will fuel growth throughout 2013 for Ericsson. LATAM Mobile SPs have already selected key LTE suppliers, with Ericsson winning a majority market share. Ericsson has also demonstrated the world’s first end-to-end LTE broadcast video solution and has commercial deployment endorsements from Verizon Wireless and Telstra. Ericsson's unique combination eMBMS, HEVC and MPEG DASH, three new standards, enables Mobile SPs to provide premium video services with guaranteed quality and cost-efficient delivery over LTE.

Market Trends/Predictions

ACG sees several trends emerging in Mobile SP CapEx outlays. 3G network CapEx in RAN and core segments has eroded, and ACG expects a decline of 20–30% Y/Y through 2013 across many regions. Mobile IP Backhaul CapEx will continue to grow in double digits through 2013 globally as operators continue to optimize cell site capacity and network operations costs. ACG expects CapEx spend to increase 15–20% Y/Y on mmW and NLOS technologies as these are optimal for small cells and provide deployment flexibility in metro zones where macro cell sites do not make economic sense. ACG expects EPC to grow at record pace, averaging 45–50% Y/Y in 2013 with the US and Canada undergoing national builds, and with APAC and LATAM driving additional LTE ‘deployment’ revenues.

Next Wave of Mega LTE Networks: India & China

For the LTE industry overall, ACG predicts the next wave of growth opportunities will come from TD-LTE builds in India and China. Although the timing of mass LTE deployments in India is questionable, as policy/licensing issues, as well as regulator/tax collector actions against the country’s top Mobile SPs continue to escalate and cause unfavorable investment climate. In contrast, 3G subscriptions in China continue to skyrocket, and demand for TD-LTE networks with its inherent cost/spectrum efficiencies is at its peak. For example, China Mobile will spend more than $7 billion in CapEx this year on its TD-LTE network build. China Unicom and China Telecom will make decisions regarding TD-LTE in mid 2013, with rollouts beginning as early as end of 2013. Tier 1 vendors such as Alcatel-Lucent and Nokia Siemens are tripling investments and resources in China to support accelerated TD-LTE build-outs.

For more information about ACG's Mobile IP Infrastructure services, contact sales@acgresesearch.net