Telia Carrier, a network services provider based in Stockholm, is looking to work with large master agents and regional partners in the U.S., targeting enterprise opportunities such as SD-WAN.
The company this week launched a partner program with the goal of raising its profile in the U.S. The channel initiative aims to help partners cross-sell a range of offerings that include internet services, Ethernet, MPLS, a public cloud gateway and SD-WAN services. Telia Carrier earlier this month released a new SD-WAN offering based on Cisco’s Viptela technology.
Rob Pulkownik, head of channel sales at Telia Carrier, said the company recently built out its internal infrastructure to work with partners, creating mechanisms to track orders, pay commissions and avoid channel conflict.
“Now that we have that in place, my plan for this year is to scale up with … two more of the large masters and then regional [agents], on a more ad hoc basis,” he said.
Telia Carrier has master agent agreements in place with AppSmart (formerly WTG), Telarus and other companies.
The channel sales effort represents a shift for Telia Carrier, which has operated primarily a wholesale player, with customers including content providers, carriers, multisystem operators and ISPs. The enterprise sector was much less of a focus. Telia Carrier has staffed eight to 10 salespeople in the U.S. market, while competitors have more than 1,000 salespeople, Pulkownik noted.
Rob Pulkownik Head of channel sales, Telia Carrier
“We are not going to ramp up a sales team like that,” he said. “We are going to rely on doing a lot of this through the channel.”
Features of Telia Carrier’s partner program include a self-service portal, which lets agents keep tabs on inventory, usage, trouble tickets, invoices, customer payments and commissions, according to the company.
Telia Carrier aims to roll out an automated deal registration system in the second quarter of this year. At the moment, deal registration is a manual process.
As network engineer skills become increasingly generalized, Cisco aims to match its certifications to the skills network engineers need in their daily lives.
Announced at Cisco Live 2019, the new Cisco certification changes rolled out on Feb. 24, 2020. Experts have touted the relevant material and the myriad topics Cisco’s certifications cover with these changes and potential benefits for network engineers. With more focus on automation and software skills and less on infrequently used coding languages, Cisco aims to spring its certification tracks forward into the new decade.
The Cisco Certified Network Associate (CCNA), Cisco Certified Network Professional (CCNP) and Cisco Certified Internetwork Expert (CCIE) certifications all expanded the breadth of topics covered, yet all shrunk in size. Cisco also introduced new DevNet certifications among the other Cisco certification changes.
How did existing Cisco certifications change?
Cisco’s standard certification tracks — CCNA, CCNP and CCIE — all added new material that aims to be more relevant to current job roles and help advance the careers of network engineers. In addition to new material, the certifications also include fewer track options than before.
Cisco Certified Network Associate. CCNA is an entry-level certification for network engineers early in their careers. Formerly, Cisco issued the Cisco Certified Entry Networking Technician (CCENT) certification, which was the step before CCNA. After CCENT, CCNA offered different certifications for various career tracks, including CCNA Routing and Switching and CCNA Collaboration.
Now, CCENT is gone, and the recent Cisco certification changes transformed the CCNA from 10 separate tracks into a single unified exam, apart from the CCNA CyberOps track. Cisco author Wendell Odom said most topics in the new CCNA exam come from the former CCNA Routing and Switching track, with about one-third of new material.
A CCNA certification isn’t a prerequisite for higher certifications, yet it provides fundamental networking skills that network engineers require for current job roles.
Cisco Certified Network Professional. CCNP is an intermediate-level certification and a step up from CCNA. Similar to the CCNA changes, Cisco consolidated the CCNP certification tracks, although less drastically than with CCNA. Cisco cut CCNP from eight to five tracks, which, like CCNA, reflect holistic industry changes to bring more relevant material to Cisco’s certifications.
According to Cisco, the new CCNP tracks — which are also the new CCIE tracks — are the following:
While these are the five core exams a network engineer can take, they must also take a concentration exam within the core topic to attain a CCNP certification. If a person solely takes the core exam and passes, she receives a Cisco Certified Specialist certification in that topic area.
Network engineers can take several core or concentration exams and receive a Cisco Certified Specialist certification upon passing, which can prove to employers the engineer has those specific skills.
Cisco Certified Internetwork Expert. CCIE is an expert-level certification and a step up from CCNP. The CCIE and CCNP tracks fall under the same umbrellas and shrunk to the aforementioned five tracks. To become CCIE-certified, network engineers must take and pass one core exam — Enterprise, Security, etc. — and that topic’s corresponding lab.
Formerly, CCIE exams focused more on highly advanced skills and less on critical knowledge in areas such as network design skills. After the Cisco certification changes, the CCIE exams now include more practical knowledge for advanced network engineers.
What are the new Cisco certifications?
In Cisco’s new DevNet track, the company added three certifications that reflect the certification pyramid for standard Cisco certifications. The DevNet certifications are the following:
Cisco Certified DevNet Associate
Cisco Certified DevNet Specialist
Cisco Certified DevNet Professional
The DevNet tracks encompass network automation, software and programmability skills that Cisco certifications previously lacked and that the industry has deemed increasingly important.
While DevNet lacks a CCIE-equivalent track, the requirements for a DevNet certification reflect those of its equivalent in Cisco’s standard certifications. For example, a person must pass one core and one concentration exam to receive a Cisco Certified DevNet Professional certification.
The DevNet track’s goal is to give network engineers a certification path for skills the industry says they need and help them adapt to newer, advanced technologies — such as network automation — that employers increasingly seek out. And, as the industry continues to change, so will Cisco’s certifications.
For the Cisco Certified Network Professional and Cisco Certified Internetwork Expert — CCNP and CCIE, respectively — the Cisco certification changes mean less time flaunting advanced networking tricks and more time learning material relevant to the current job market.
Cisco announced these certification changes at Cisco Live 2019, where thesignificant cuts to the Cisco Certified Network Associate(CCNA) track garnered much attention. However, the Cisco certification changes also affected the CCNP and CCIE tracks, such as shrinking the former eight-track CCNP options to five tracks. Authors Brad Edgeworth and Jason Gooley said they believe these changes will greatly benefit CCNP and CCIE hopefuls, as the changes reflect shifts in the networking industry and network engineer job roles.
The effects of the Cisco certification changes are reflected in the new book from authors Ramiro Garza Rios, David Hucaby, Edgeworth and Gooley —CCNP and CCIE ENCOR 350-401 Official Cert Guide—which is available now. The book explores the new CCNP and CCIE Enterprise tracks that include relevant information for enterprise network engineers.
Editor’s note:The following interview was edited for length and clarity.
How have the Cisco certification changes affected CCNP and CCIE? What remains the same?
Brad Edgeworth: [Cisco] is adding more width to the knowledge required. It’s adding more programmability and automation, because that’s becoming more integrated into teams. Also, it’s trying to take advantage of more virtualized platforms.
Jason Gooley: The certifications are becoming more streamlined. They’re modular, so you can pick the technology core, then focus on a specialty and become certified in that direction. In addition, newer technologies such as software-defined access [SD-Access] or software-defined WAN [SD-WAN] are part of these exams.
Not a lot was removed. The level of knowledge you had to know before has grown, because we include what was there before and add a bunch of new technologies.
Edgeworth: Cisco is going back to what is relevant to jobs. Some technologies that are not as common, like frame relay, were removed. The core fundamentals of networking still reside within the certification exams, and Cisco built on top of them.
Gooley: You have to know what was asked [in the exams] before in addition to these new technologies. That fits with what customers see in work environments now. You’re certified in what you see in the industry versus an exam with some technology you might not use. It’s structured around current job roles.
When I took the CCNP, there were four exams. Now, you can take two — technology core and concentration — and become CCNP certified. The structure completely changed, which I think is for the better. As far as technology, things like SD-WAN, SD-Access and programmability become more robust because that’s what customers and the industry are leading to.
Edgeworth: The CCNP Routing and Switching exam before was great but never took wireless into account, which is what most enterprise customers use. Now, that’s integrated into it.
With CCIE, it used to be: What router ninja tricks can you do? CCIEs would maybe not have fundamentals for network design, so network design was integrated as a component of the CCIE practical exam. Design concepts have become a core specialization with CCNP, as well.
Where do you see Cisco certifications and the industry heading in the next 20 years?
Edgeworth: In the industry, there will be more automation and businesses becoming more digital. Another big thing is security. How do you integrate security throughout the service? The industry lagged with that. There’s going to be more automation and security integration for dotting i’s and crossing t’s to make sure data is correct and maintains its privacy.
Gooley: As job roles change and customers adopt different technologies, the certifications will follow. As the certifications evolve over time, they’ll follow what’s in the industry and what customers go through. That’s why we didn’t remove a lot from the certifications, because it’s still out there.
For Cisco to redo the entire certification program, as well as introduce a new line that focuses specifically on automation, software and programmability skills — that’s in response to the industry, and that’s critical. When you evolve your skill set and move toward newer technologies and automation, you still need to know how it works before you automate it. You can automate failure as fast as you automate success.
Edgeworth: You have to have fundamentals because of what you automate. Learn the trade, not tips of the trade, because tips of the trade come from learning the trade.
Going after a certification is nice. Obtaining the certification is nicer. But failure is part of the process. Learning on the journey is critical. I wouldn’t be where I am today if I hadn’t failed. [The first time] I tried for CCIE, I failed. But the knowledge I gained during the process allowed me to enter other opportunities to grow my career. While succeeding is nice, it’s about the knowledge you gain on the way.
Gooley: I went for the CCDE [Cisco Certified Design Expert] three times, and I still haven’t passed. You learn the technology and best practices in going for it. Even if you don’t pass, you’ve still enhanced your skill set, and it’s valuable. Everybody eventually has to get up and dust themselves off.
What’s nice about social media and the community is when you fail, you’re held accountable when you say it. Then other people come out of the woodwork saying you’re not alone. That helps everybody learn together. Embrace the journey. The journey is where you learn everything and have the fun.
Although the Cisco Certified Network Professional track no longer has prerequisite exams, most CCNP exams still require an understanding of the networking topics found in the reworked Cisco Certified Network Associate, or CCNA, exam.
For the CCNP and Cisco Certified Internetwork Expert (CCIE) ENCOR 350-401 exam, a significant portion of the material includes information from the new CCNA exam. Authors Ramiro Garza Rios, David Hucaby, Brad Edgeworth and Jason Gooley cover both the old and new material in their guidebook CCNP and CCIE ENCOR 350-401 Official Cert Guide, which is available now.
The ENCOR 350-401 exam — which stands for Enterprise Core — particularly emphasizes Cisco’s move from uncommon, advanced capabilities to the networking requirements for current job roles.
Below is an excerpt from the guide: Chapter 6, “IP Routing Essentials.” This chapter covers fundamental routing protocols — many of which have remained from when the authors themselves began to study for Cisco certification exams.
When Edgeworth first studied for the Cisco certification exams, he said understanding how routers think and operate was the most challenging part. As an author, he has tried to write chapters in a way that provides in-depth perspective, yet also shows how technologies and protocols work within configurations. Edgeworth suggested CCNP and CCIE ENCOR 350-401 hopefuls participate in labs to put the concepts they learn from books into practice.
Gooley, on the other hand, found unstructured, solo studying the most challenging, saying he felt alone when he first started studying for his Cisco certifications. He suggested that hopefuls should lean on the community, whether that’s in person or through social media. People can hold each other accountable for studying and readers can reach out to the authors themselves if they have questions.
In addition to potentially challenging new topics in the CCNP and CCIE ENCOR 350-401 exam — such as programmability and software-defined WAN — Edgeworth and Gooley said they are pleased with how relevant the ENCOR 350-401 exam is to current job roles. CCNP and CCIE hopefuls can expect to learn and solidify skills they use daily at their jobs, including the IP routing fundamentals.
Edgeworth said this chapter covers many routing essentials, such as Open Shortest Path First (OSPF) and Border Gateway Protocol (BGP) — topics the new CCNA exam also includes. The chapter delves into fundamental knowledge network engineers need for jobs and explores routing essentials in a vendor-agnostic way, as OSPF is OSPF and BGP is BGP regardless of which vendor platform an engineer uses, according to Edgeworth.
While Cisco’s updated Cisco Certified Network Associate — or CCNA — certification track shrunk to a single path and single exam, CCNA hopefuls must know a broad range of both networking basics and emerging networking technologies in order to pass the exam.
Cisco announced sweeping changes to its certification tracks in June 2019, and the newCCNAexam derives from one of the largest changes in Cisco history, according to Cisco author Wendell Odom. Odom, author of every CCNA Official Cert Guide, wrote two new volumes of his guides for the CCNA 200-301 exam. The singular path of the new CCNA exam is smaller overall compared to past exam versions, yet the extensive amount of material — both old and new — necessitated two volumes.
Both Volumes 1 and 2 cover various traditional networking topics, such as virtual LANs (VLANs) and basic IP services, as well as newer networking technologies, such as network automation. Odom said the new CCNA exam includes a lot for engineers to learn but also contains relevant and useful material for thecurrent job market.
Editor’s note: The following interview was edited for length and clarity.
Can you compare details of the former and the new CCNA exams?
Wendell Odom: If you took the old CCNA Routing and Switching exam blueprint, about half those topics are in the new CCNA exam. The literal words are there. It’s not just the same topic — it’s copied-and-pasted topics from the old to the new.
Then, the new exam has topics that weren’t in any of the old. It has a few you might say came from CCNA Collaboration or CCNA Data Center. For the most part, the new topics [show] the world is changing and IT changes quickly. These are new things Cisco finds important for routing and switching, like automation and cloud. Now, it introduces intent-based networking to CCNA for the first time.
If you view the old as 100 points in volume, the new is about 75% of that — 75 points. Fifty points are old exam topics that stuck around: VLANs, VLAN trunks, IPv4 and IPv6 routing, Layer 3 filters, sub-Layer 2 filtering with port security, security protocols, basic IP services, like SNMP [Simple Network Management Protocol] and NTP [Network Time Protocol].
Click to learn more about this book.
Now, there’s more OSPF [Open Shortest Path First] — particularly, OSPF network types. On an Ethernet interface, you’ve got two or more routers that run OSPF connected to the same Ethernet. They elect a designated router, which causes OSPF to model the connected subnet differently. It changes OSPF operation on that LAN.
That’s typical on a LAN, but if you use Ethernet in WANs — particularly point-to-point WAN links — you don’t want LAN-like OSPF behavior electing a designated router. To change that, in Cisco routers, you change the OSPF network type to point-to-point instead of the default broadcast type, which is what causes it to act like a LAN.
The new Volume 1 has four chapters on wireless LANs. It’s basic: What’s an access point [AP]? What are the different wireless standards? How would you configure an AP to be a stand-alone AP? How would you do it with a wireless LAN controller? To a networker, it’s not very deep, but it’s your first step, and there’s a lot in CCNA that are first steps in learning technologies.
Now, there’s DHCP [Dynamic Host Configuration Protocol] snooping and dynamic ARP [Address Resolution Protocol] inspection. And the new CCNA exam mentions TFTP [Trivial File Transfer Protocol] and FTP specifically.
The old had basics of what I call ‘controller-based networking;’ there’s more now. It talks about underlays and overlays, which now gets you ready for software-defined access. The old and new CCNA exams have a lot about the old way to do LANs — how you build switch networks, Spanning Tree Protocol, etc.
If you studied now for everything except newer technologies, which is 10% of the exam blueprint, it’d seem like traditional networking technology. Then, you get into newer, evolving technologies. Now, we’re pushing the baby birds out of the nest because … you’re going to get a lot of this in the CCNP Enterprise Core, etc. I’m glad some of it is in CCNA.
What questions have you gotten about the new CCNA exam?
Odom: Oddly enough, there’s not much worry about new topics. ‘Do I need to know Python?’ That’s probably most common because exam topics don’t mention Python. You think automation, and you think your first step is a programming language. You can actually learn everything in CCNA for automation without knowing Python.
People quickly zero in on technical questions: Layer 2, Layer 3 interactions. People get confused about encapsulation. OSPF concepts are more common — typically, LSAs [link-state advertisement], what those mean and whether that’s important. ‘Do I need to understand what a Type 1, Type 2 and Type 3 LSA is?’ I don’t know how important that is for the exam depending on the version. But if you’re going to use OSPF, you need to know what it is for real life.
I’m happy with how [the new CCNA exam] balances newer automation features and technologies — not overwhelming newbies with too much new and giving the foundation they need to get a real job. I think Cisco hit the right balance. People will enjoy the topics they learn, both for learning and for how it matches real jobs today. Cisco did this particular exam right.
Many organizations rely on a virtual private network, particularly those with a large number of remote workers who need access to resources.
While there are numerous vendors selling their VPN products in the IT market, Windows administrators also have the option to use the built-in VPN that comes with Windows Server. One of the benefits of using Windows Server 2019 VPN technology is there is no additional cost to your organizations once youpurchase the license.
Another perk with using a Windows Server 2019 VPN is the integration of the VPN with the server operating system reduces the number of infrastructure components that can break. An organization that uses a third-party VPN product will have an additional hoop the IT staff must jump through if remote users can’t connect to the VPN and lose access to network resources they need to do their jobs.
One relatively new feature in Windows Server 2019 VPN functionality is the Always On VPN, which some users in various message boards and blogs have speculated will eventually replace DirectAccess, which remains supported inWindows Server 2019. Microsoft cites several advantages of Always On VPN, including granular app- and traffic-based rules to restrict network access, support for both RSA and elliptic curve cryptography algorithms, and nativeExtensible Authentication Protocolsupport to enable the use of a wider variety of advanced authentication methods.
Microsoftdocumentation recommendsorganizations that currently use DirectAccess to check Always On VPN functionality before migrating their remote access processes.
The following transcript for the video tutorial by contributor Brien Posey explains how to install the Windows Server 2019 VPN role.
In this video, I want to show you how to configure Windows Server 2019 to act as a VPN server.
Right now, I’m logged into a domain joined Windows Server 2019 machine and I’ll get the Server Manager open so let’s go ahead and get started.
The first thing that I’m going to do is click on Manage and then I’ll click on Add Roles and Features.
This is going to launch the Add Roles and Features wizard.
I’ll go ahead and click Next on the Before you begin screen.
For the installation type, I’m going to choose Role-based or feature-based installation and click Next. From there I’m going to make sure that my local server is selected. I’ll click Next.
Now I’m prompted to choose the server role that I want to deploy. You’ll notice that right here we have Remote Access. I’ll go ahead and select that now. Incidentally, in the past, this was listed as Routing andRemote Access, but now it’s just listed as a Remote Access. I’ll go ahead and click Next.
I don’t need to install any additional feature, so I’ll click Next again, and I’ll click Next [again].
Now I’m prompted to choose the Role Services that I want to install. In this case, my goal is to turn the server into a VPN, so I’m going to choose DirectAccess and VPN (RAS).
There are some additional features that are going to need to be installed to meet the various dependencies, so I’ll click Add Features and then I’ll click Next. I’ll click Next again, and I’ll click Next [again].
I’m taken to a confirmation screen where I can make sure that all of the necessary components are listed. Everything seems to be fine here, so I’ll click Install and the installation process begins.
So, after a few minutes the installation process completes. I’ll go ahead and close this out and then I’ll click on the Notifications icon. We can see that some post-deployment configuration is required. I’m going to click on the Open the Getting StartedWizard link.
I’m taken into the Configure Remote Access wizard and you’ll notice that we have three choices here: Deploy both DirectAccess and VPN, Deploy DirectAccess Only and Deploy VPN Only. I’m going to opt to Deploy VPN Only, so I’ll click on that option.
I’m taken into the Routing and Remote Access console. Here you can see our VPN server. The red icon indicates that it hasn’t yet been configured. I’m going to right-click on the VPN server and choose the Configure and Enable Routing and Remote Access option. This is going to open up the Routing and Remote Access Server Setup Wizard. I’ll go ahead and click Next.
I’m asked how I want to configure the server. You’ll notice that the very first option on the list is Remote access dial-up or VPN. That’s the option that I want to use, so I’m just going to click Next since it’s already selected.
I’m prompted to choose my connections that I want to use. Rather than using dial-up, I’m just going to use VPN, so I’ll select the VPN checkbox and click Next.
The next thing that I have to do is tell Windows which interface connects to the internet. In my case it’s this first interface, so I’m going to select that and click Next.
I have to choose how I want IP addresses to be assigned to remote clients. I want those addresses to be assigned automatically, so I’m going to make sure Automatically is selected and click Next.
The next prompt asks me if I want to use a RADIUS server for authentication. I don’t have a RADIUS server in my own organization, so I’m going to choose the option No, use Routing and Remote Access to authenticate connection requests instead. That’s selected by default, so I can simply click Next.
I’m taken to a summary screen where I have the chance to review all of the settings that I’ve enabled. If I scroll through this, everything appears to be correct. I’ll go ahead and click Finish.
You can see that the Routing and Remote Access service is starting and so now my VPN server has been enabled.
N54L only ever used as a backup machine. Stuck a network card in to allow it to run xpenology latest variant. 4gb memory upgrade. Boxed with keys etc. I’m in North London (N10), so if you can collect I would do for £70. Please let me know. Cheers.
Working on Microsoft Azure platform, Mohanty and his colleagues used a Convolutional Neural Network model to come up with a solution that can identify and count penguins with a high degree of accuracy. The model can potentially help researchers speed up their studies around the status of penguin populations.
The team is now working on the classification, identification and counting of other species using similar deep learning techniques.
Building AI to save the planet
A long-time Microsoft partner headquartered in Hyderabad in India, Gramener is not new to leveraging AI for social good using Microsoft Azure. It was one of the earliest partners for Microsoft’s AI for Earth program announced in 2017.
“I believe that AI can help make the world a better place by accelerating biodiversity conservation and help solve the biggest environmental challenges we face today. When we came to know about Microsoft’s AI for Earth program over two years ago, we reached out to Microsoft as we wanted to find ways to partner and help with our expertise,” says Kesari.
While the program was still in its infancy, the teams from Gramener and Microsoft worked jointly to come up with quick projects to showcase what’s possible with AI and inspire those out there in the field. They started with a proof of concept for identifying flora and fauna species in a photograph.
“We worked more like an experimentation arm working with the team led by Lucas Joppa (Microsoft’s Chief Environmental Officer, and founder of AI for Earth). We built a model, using data available from iNaturalist, that could classify thousands of different species with 80 percent accuracy,” Kesari reveals.
Another proof of concept revolved around camera traps that are used for biodiversity studies in forests. The camera traps take multiple images whenever they detect motion, which leads to a large number of photos that had to be scanned manually.
“Most camera trap photos are blank as they don’t have any animal in the frame. Even in the frames that do, often the animal is too close to be identified or the photo is blurry,” says Mohanty, who also leads the AI for Earth partnership from Gramener.
The team came up with a two-step solution that first weeds out unusable images and then uses a deep learning model to classify images that have an animal in them. This solution too was converted by the Microsoft team into what is now the Camera Trap API that AI for Earth grantees or anyone can freely use.
“AI is critical to conservation because we simply don’t have time to wait for humans to annotate millions of images before we can answer wildlife population questions. For the same reason, we need to rapidly prototype AI applications for conservation, and it’s been fantastic to have Gramener on board as our ‘advanced development team’,” says Dan Morris, principal scientist and program director for Microsoft’s AI for Earth program.
Anticipating the needs of grantees, Gramener and Microsoft have also worked on creating other APIs, like the Land Cover Mapping API that leverages machine learning to provide high-resolution land cover information. These APIs are now part of the public technical resources available for AI for Earth grantees or anyone to use, to accelerate their projects without having to build the base model themselves.
Startups played a pivotal role in disrupting the business of network switching. Today, they’re on track to do the same to routing.
Software under development by upstarts Arrcus, DriveNets and Volta Networks represents a new routing architecture, industry analysts agreed. Cloud service providers, SaaS providers, telcos and the largest financial institutions are the most likely candidates for deploying the networking startups’ technology in the data center and at the edge.
The vendors’ software could also be useful for peering among internet service providers and for data center interconnects (DCIs). Colocation companies like Equinix, Digital Realty Trust and Global Switch use DCIs to connect their facilities to customer data centers.
Market research firm IDC recently named the three companies 2019 innovators for their work in decoupling routing software from its underlying hardware. Separating management, control and data planes from the device make it possible to run the software on commodity products powered by merchant silicon from companies like Broadcom and Intel.
Severing software from hardware and running it on commodity gear — a process called disaggregation — reduces operational expenses. Companies can lower labor costs by managing multiple routers at once, instead of each one separately. The architecture also adds flexibility by making it possible to distribute and manage physical and virtual routers across data centers or at the network edge.
“Effectively, you’ve got a Lego that you can mix and match based on your requirements,” said Brad Casemore, an analyst at IDC. “It leads to a standardized environment where you can run the same software across all of it.”
Disaggregation from switching to routing
Disaggregation in network switching, a nearly 10-year trend, forced incumbents Cisco and Juniper Networks to acquire startups that had developed software capable of providing centralized network management. The transition led to an overhaul in the way the companies’ products manage switching fabrics.
New technologies developed by Arrcus, DriveNets and Volta show that there’s “an evolution in disaggregation to the routing layer,” Casemore said. Each of the vendors is initially targeting their products at communication and cloud service providers.
Here is a brief description of each of the networking startups, including the key differentiators and market challenges listed in the 2019 IDC Innovators report on disaggregated routing platforms:
— Arrcus built a network operating system, called ArcOS, with extensive routing protocol support. This year, for example, the vendor incorporated the Link State Vector Routing (LSVR) protocol into ArcOS for organizations running hyperscale data centers and large cloud environments.
Arrcus has built its data plane adaptation layer to separate ArcOS from the underlying hardware. ArcOS is also the first independent NOS to support devices powered by Broadcom’s Trident 3, Tomahawk 3, Jericho+ or Jericho2 network silicon. The Jericho2 platform is for 100 Gb and 400 Gb routing.
Despite its innovative technology, Arrcus still has to prove it can deliver significant cost savings and ROI. The company also has to show a simple process for buying and supporting the underlying hardware.
Arrcus, based in San Jose, Calif., has more than 60 employees and has raised $45 million in funding.
— DriveNets developed a container-based router control plane for merchant silicon-based white boxes. Hardware manufacturers bundle the software with their products and sell it under a license that is free from capacity constraints.
The architecture provides carriers with a routing model that uses a cluster of low-cost white boxes capable of scaling to any size. DriveNets based the model on the one used in hyperscale data centers.
DriveNets’ hurdles include convincing communication service providers to change how they procure, deploy and manage router infrastructure. “The adoption of the DriveNets architecture might be slowed by the need for communication service providers to redesign internal processes and management systems,” IDC said.
DriveNets, based in Ra’anana, Israel, has more than 200 employees and has raised $117 million in funding.
— Volta built a cloud-native, cloud-hosted control plane that can spin up and manage as many as 255 instances of virtual routers on a single, on-premises commodity switch. The use of switching gear provides a “significant cost advantage,” while also making Volta technology useful for provider edge routing. Volta’s technology could be helpful to carriers overhauling cell sites to support next-generation 5G wireless technology.
Volta’s technology and its subscription model that covers support, maintenance and hardware warranty could provide significantly lower capital and operational expenses. However, as a startup, in a competitive industry, it faces a “significant challenge” in winning deals over better-known competitors with more money.
Volta, based in Cambridge, Mass., has 51 employees and has raised $3.3 million in funding.
Moving toward software-based routing
Companies with hyperscale data centers, like Amazon, Facebook, Google and Microsoft, have favored disaggregated networking software on standardized hardware for years. Today, major service providers and financial institutions use the same white box switches. Users include AT&T, Comcast, Verizon, JPMorgan Chase and Fidelity Investments.
As a result, in 2018, the share of the global Ethernet data center switching market held by Cisco and Juniper fell, while that of bare-metal switching manufacturers increased, according to IDC.
Analysts believe the same dynamics will likely play out in routing. “People are now noticing and realizing that white box approaches can work. They’re mature,” said Roy Chua, a principal analyst at AvidThink.
Brad CasemoreAnalyst, IDC
Analysts expect carriers to seriously consider white box routers as they build out their network edge to deliver 5G services.
“They’re actually trying to move away from [physical] routers and toward software-based routing,” said Lee Doyle, principal analyst at Doyle Research. “None of this has been hugely deployed yet. But I think we’re going to see significant deployments in 2020 and 2021 in the 5G market.”
Routing sales for Cisco and Juniper have been declining. However, the decrease is primarily due to carriers cutting back on spending after they found they couldn’t wring any more revenue from consumers, Casemore said.
But with 5G deployments on the horizon, incumbents like Cisco and Juniper are probably watching networking startups closely to see which ones are winning deals for routing technology.
“Potentially, these companies become M&A targets if they have traction in some high-value accounts,” Casemore said.
Google has introduced several network monitoring tools to help companies pinpoint problems that could impact applications running on the Google Cloud Platform.
Google launched this week the first four modules of an online console called the Network Intelligence Center. The components for monitoring a Google cloud network include a network topology map, connectivity tests, a performance dashboard, and firewall metrics and insights. The first two are in beta, and the rest are in alpha, which means they are still in the early stages of development.
— Google is providing Google Cloud Platform (GCP) subscribers with a graphical view of their network topology. The visualization shows how traffic is flowing between private data centers, load balancers, and applications running on computing environments within GCP. Companies can drill down on each element of the topology map to verify policies or identify and troubleshoot problems. They can also review changes in the network over the last six weeks.
— The testing module lets companies diagnose problems with network connections within GCP or from GCP to an IP address in a private data center or another cloud provider. Along with checking links, companies can test the impact of network configuration changes to reduce the chance of an outage.
–The performance dashboard provides a current view of packet loss and latency between applications running on virtual machines. Google said the tool would help IT teams determine quickly whether a packet problem is in the network or an app.
–The firewall metrics component offers a view of rules that govern the security software. The module is designed to help companies optimize the use of firewalls in a Google cloud network.
Getting access to the performance dashboard and firewall metrics requires a GCP subscriber to sign up as an alpha customer. Google will incorporate the tools into the Network Intelligence Center once they reach the beta level.