As Principal Analyst of Transport and Routing Infrastructure, Glen analyzes technology, product, and partnership initiatives of vendors who supply carrier infrastructure equipment. Specifically, focusing on vendors that produce core routers, edge switching, optical transport, data center interconnection, mobile backhaul, network management and operational support systems.
• Segment Routing Deployed in a Backbone Network: Cisco and China Unicom collaboration, a two year endeavor, enables the operator’s nationwide MPLS VPN network to support on-demand services. Cisco notes that this represents China’s first Segment Routing (SR) deployment and leverages its SDN technologies to provide an end-to-end solution. SR also helps mitigate the use of more complicated IP protocols (like RSVP-TE), which is a big advantage and leverages the power of SDN’s centralized management model. .
• Advanced Vendor Services Accelerated Delivery: Cisco’s Advanced Services teams provided full life cycle support and coordination with China Unicom and third-party suppliers, and tackled SDN, NFV, cloud computing, ultra-broadband networking, and related technology issues for the operator. The use of vendor supplied services has always been part of the “vendor – operator” model, but virtualized technologies pose additional time-to-market issues such as mastering new technologies and dealing with the complexity of integrating multiple components.
The announcement by Cisco and China Unicom represents a tangible example of how current infrastructure networks can be transformed to deliver flexible cloud-based services by exploiting new routing and networking techniques, such as segment routing which aligns well with centralized management and control (i.e., SDN) paradigms. What makes the solution significant is the combination of multiple technologies, SR, SDN controllers, service orchestration, and the ability to quickly compute optimal traffic paths through an IP network using a path computation engine (PCE). Individual components on their own cannot deliver the resultant end-to-end solution, which supports China Unicom’s ability to offer a range of named cloud-based services (Cloud Network Connection, Cloud Networking, Cloud Broadband, Unicom Cloud Shield, Intelligent Boutique Video Network, and Boutique Financial Network) which it noted in a joint press release with Cisco. Continue reading “China Unicom and Cisco Collaborate to Deliver Cloud + Network Capabilities”→
• Automates Complex Network Slicing Process: The Contrail Network Slicing Bot aims to simplify the highly complex process of defining, creating, provisioning and managing network slicing across both physical and virtual infrastructures. Network slicing is a fundamental tenet of all 5G architectures.
• Addresses Skill Set Shortage: Juniper’s network Bots, including the Slicing Bot, help alleviate the lack of internal staff education and higher level skillsets required to use network automation tools, which are often complex and require deep network knowledge. The Slicing Bot uses a high level human friendly language to translate operator intent into actionable workflows.
Juniper’s “Slicing Bot” launch enables operators to leverage technologies such as machine learning, network telemetry and SDN control, to help reduce the complexity they face as they attempt to deploy virtual networks to meet the growing demand for faster time-to-service. The new Contrail Slicing Bot joins three other Bots, which the company launched in December 2017. The Bots enable operators to operate at a business level (i.e., intent) to help automate complex network functions that are often error prone and time consuming due to manual processes and the lack of operational skills. The three initial Bot applications offered by Juniper included PeerBot, TestBot and HealthBot, which address key pain points for operators; the new Slicing Bot tackles the highly complex, but essential process of carving up the network into virtual slices, a fundamental building block for supporting 5G services. Continue reading “MWC18: Juniper Expands its “Bots” Portfolio with the Contrail Slicing Bot, Helping to Automate More Complexity”→
• Webscale influence on new product developments is being felt in data center interconnect and massively scalable switching and transport gear.
• Traditional telcos, however, are using these new Webscale-driven platforms to retool their own networks and prepare to deliver more agile services to protect their service base.
Since the first public networks were built, incumbent telcos like Vodafone, Deutsche Telekom and AT&T have ruled network equipment roadmaps and investments – dictating the features, operation and capabilities of new networking products. However, more recently Webscale operators are the influencers. Whereas vendors had been focused on providing the high capacity and broad coverage required by rapid smartphone adoption and a massive increase in video traffic, the focus has steadily turned to support massive, low-latency throughput between the data centers that serve the most popular brands on the Internet, dubbed FAMGA (Facebook, Apple, Microsoft, Google and Amazon). Traditional network operators are scrambling to adapt to this change in service delivery focus, and better prepare their networks for the emerging 5G and Internet of Things (IoT) era. Continue reading “Webscales Yield Heavy Market Influence: Google, Facebook and Others Driving Network Equipment Roadmaps”→
Introduces Slicing for Backhaul: The ZXCTN 609 supports separate backhaul slices (tunnels), each with independent performance characteristics, meeting 5G demands for low-latency, high-speed, flexible connections.
100G Backhaul Link Support: The ZXCTN 609 expands ZTE’s Flexhaul series to support 100G links to handle expected high-bandwidth 5G backhaul speeds, with high-availability features such as protection switchover and SDN control.
ZTE leveraged this year’s MWC Shanghai 2017 to further its stake in the emerging 5G infrastructure market by expanding its Flexhaul series backhaul platform to support 100G links. 100G is needed to cope with the massive capacity requirements expected as 5G comes to life in the next few years. The ZXCTN 609 also includes the company’s FlexE tunnel technology, which it announced as part of the ZXCTN 6180H launch, bringing the equivalent of network slicing to the backhaul network. ‘Flexibility’ is clearly the focus, with FlexE supporting a variety of service characteristics for applications such as enhanced mobile broadband (eMBB), massive machine type communications (mMTC), and ultra-reliable and low-latency communications (uRLLC), which have vastly different transport requirements. ZTE has collaborated closely with multiple operators to craft mobile network solutions that meet a range of application types and capacities. In addition to being visible in China Mobile and Ncell Axiata, ZTE successfully completed tests in seven major scenarios that are part of the second phase of China’s national 5G tests and set multiple records for network speeds and performance. Continue reading “MWC Shanghai 2017: ZTE Addresses ‘5G Network Slicing Backhaul’ Requirements with New Solution”→
• The telco cloud requires an IP and optical core that can handle massive traffic growth and data center interconnection requirements. Huawei’s Network Cloud Engine (NCE) solution brings a cloud operational model to the network core.
• Huawei’s NetEngine 9000 (NE9000) IP core router delivers leading capacity to address massive 100G-centric connectivity requirements, from drivers such as 4K video.
This year, at Huawei’s 14th Annual Analyst Summit, affectionately referred to as HAS2017, we found a strong focus on the infrastructure layers of the network – specifically the IP core and optical core. The attention given these network layers has historically been sporadic, driven by necessary upgrades in link and/or switching capacity to keep up with growing traffic demands, a cycle that has occurred in four to eight year cycles, and further paced by technological breakthroughs. Recent industry discussion related to telco cloud seemed to be focused on network functions virtualization (NFV) in an effort to create an environment capable of supporting traditional carrier connectivity businesses and counter the threat from OTT players. However, with the arrival of web-scale providers and moves by operators to adopt a more flexible data center-centric architecture, attention is now focused on optimizing the core layers of the network.
With respect to SDN, the IP and optical domains have evolved separately, each having developed software to enable programmable control and management for the elements in their respective domains. To establish a common solution platform, Huawei introduced its Network Cloud Engine (NCE) to provide an integrated end-to-end solution for each domain, as well as provide the ability to combine multiple domains and multiple layers. In addition to the IP and optical domains, NCE includes additional solutions designed to address the access, metro and other network layers (to be covered in separate reports and blogs). Huawei has also expanded its flagship Network Engine 9000 (NE9000) core router portfolio and provided an update on its progress and new capabilities. The idea of providing a common control and management model for a multi-layer multi-level core infrastructure is not new, competitors are also pursuing similar approached to optimize the performance and efficiency of their core infrastructures.
CloudBackbone: The IP core layer, under Huawei’s NCE architecture, is controlled by the CloudBackbone solution, which includes the company’s Agile controller for IP, network service orchestration and a suite of common management functions. CloudBackbone provides support to address services such as HD video and features like automated service provisioning and network security. Also included is the ability to optimize traffic across the optical layer, which promises to deliver significantly greater bandwidth efficiency, support traffic on demand and improve provisioning efficiency.
CloudOptiX: The optical layer, under Huawei’s NCE architecture, is controlled by the CloudOptiX solution, which provides the equivalent management for the core optical transport layer as CloudBackbone does for the IP layer. CloudOptiX leverages Huawei’s Agile TSDN controller as well as common management functions and orchestration. By integrating IP + Optical, Huawei projects a 40% TCO savings through multi-layer planning and improvements in reliability with multi-layer restoration.
IP Core Advances: Along with Huawei’s NCE launch, the vendor provided an update on its flagship NetEngine 9000 (NE9000) IP core router portfolio which addresses massively scalable data center-centric interconnection (DCI) requirements, as well as traditional and emerging IP core routing functionality. The NE9000 (initially launched at HAS2015 in a 20-slot version) now includes a smaller 8-slot version, targeted to smaller core applications. Both models provide massive low latency 100G connectivity with port densities up to 400/160 per chassis, respectively. The design provides improved power/space efficiency through the use of advanced thermal techniques and expanded network programmability via Huawei’s programmable “Solar 5.0” silicon. Huawei further noted that the NE9000 is now deployed in ten service provider networks, and the NE9000-8 model will be commercially available in October 2017, along with 4 Tbps per slot line cards, to further increase capacity and density. The ability to support network slicing is also included, which enables operators to leverage one platform to support fixed, mobile and enterprise core applications.
The NCS and new NE9000 routing capacities highlighted at HAS2017, help us to remember that behind the software, is a highly capable data plane to deliver on the promise of next generation services. Capabilities such as Huawei’s NCE are well timed to support carrier transition to a DCI to cope with increased east – west traffic patterns, expected to grow >30% CAGR; support the adoption of fewer, but larger data centers, located outside of populated areas to address factors such as facility cost and power consumption; and adopt a more cloud-based operational model to automate provisioning and improve time to service. The noted capabilities also support the notion that carriers will want to manage their networks and cloud environments together, not as independent silos.
Increased east-west server-to-server data center traffic leaves traditional perimeter defense mechanisms challenged to provide adequate trust and privileges for virtual machines (VMs) while current zero-trust mechanisms consume valuable server resources.
Netronome’s new Agilio CX intelligent server adapter promises to deliver 25 Gbps of throughput; integration with Open vSwitch firewall and Mirantis OpenStack delivers the benefits of hardware acceleration and improved VM performance.
At the August 2016 OpenStack Summit, Netronome announced enhancements to its Agilio Server Networking Platform with the introduction of the Agilio CX dual-port 25GbE intelligent server adapter (ISA) and Agilio OVS Firewall software. The new platform is integrated with the Mirantis OpenStack solution, with the goal of easing cloud-based provisioning, and promises to improve performance and scale when implementing Linux Firewall-based as well as zero-trust security using OpenStack security groups. As service providers migrate their infrastructures to a data center model to meet growing demands for cloud-based services, issues such as scale and security increase ever more. Continue reading “OpenStack Summit 2016: Netronome Offers Scalable Zero-Trust Security and Higher-Performance Connectivity with New 25G Adapter”→
• Supports Both VM and Container Stacks: Cloudify 3.4 simultaneously supports VM stacks, container stacks and hybrid stacks; providing support for current VM deployments while migrating to more efficient container-based solutions.
• Leverages Open Source Innovation: Cloudify 3.4 leverages open source innovations such as Kubernetes technology, which helps support distributed container-based applications and deliver the hybrid stack capabilities.
The new Cloudify 3.4 release from GigaSpaces removes significant deployment barriers for service providers (and enterprises) that are considering migrating their virtualization architectures from a virtual machine (VM)-based model to a more streamlined and efficient container-based architecture to support microservices. The merits of this transition are covered well in many industry blogs and reports and will not be the focus here. However, what’s new is that this new version of Cloudify supports both environments (VM and container) as well as a combination of the two (hybrid). This is an important capability, because as the industry moves beyond simply proving the basics of network virtualization, traditional service provider strengths such as performance, scalability and efficiency move front and center. Cloudify 3.4 promises to support current implementations while enabling new applications to be container-based, thereby providing users a flexible migration path to the future. In addition to the dual-stack support, the new release supports fully automated “in-place” infrastructure upgrades, which helps to minimize down-time and eases migration to new software versions without impacting on-going services.