How Dyn is rising to the cloud challenge

Trying to capture an end-to-end picture of application performance across a single enterprise is challenging enough. Providing that level of visibility for hybrid- and public-cloud-enabled applications presents a whole new level of difficulty. Enter Dyn, which is an early leader in the emerging internet performance management (IPM) market.

Dyn Chairman Jim Baum, formerly CEO of analytics company Netezza, spoke recently with IDG Chief Content Officer John Gallant about Dyn’s cloud-based IPM portfolio. Baum says Dyn combines visualization and traffic steering capabilities with a slew of data about what’s happening across the global internet to help enterprises monitor and manage critical applications beyond the firewall. Baum discussed the competitive dynamics of the IPM market, and how Dyn’s tools are helping companies improve business processes and save money. He also explored how the lessons learned at Netezza are shaping Dyn’s strategy.

Your corporate vision statement says that Dyn is a company ‘connecting people, content and commerce through a single global internet’. That’s a pretty big mission. How exactly do you make that a reality?

Dyn has positioned itself in an emerging category that we refer to as IPM, or internet performance management. The strategy is to tie together people and assets – specifically cloud-oriented assets – in a way that provides a high degree of visibility, transparency, resiliency and performance. Those really are at the core of what Dyn does. We’re leveraging, particularly in the enterprise space, the broad-based migration from captive data center-oriented computing to computing as a service and cloud offerings.

Enterprises today need to connect with, obviously people: their employees, their shareholders, their customers, their suppliers, their various constituents. They need to connect devices. Think about what’s happening in IoT, the connection of an ever expanding number of devices and leveraging the data from those devices to make ever more mission-critical business decisions. They’re doing that through an increasingly new form of distributed infrastructure, largely in the form of cloud computing and CDNs [content distribution networks] for content distribution to the edge.

Dyn is focused on making all those connections work and making those connections work in a way that is performant and resilient in an environment where, frankly, enterprises no longer directly control the infrastructure. When the infrastructure is in the form of a data center or even a distributed data center, you have tremendous control over the OS levels, over the software level, over the connectivity or how things are tuned to perform. When you’re distributing these mission-critical applications and connectivity out through a global internet to every corner of the globe, you don’t control that infrastructure. Dyn is putting a control layer around that that gives people the ability to visualize, make decisions about how they deploy that infrastructure and gives them a high level of resiliency as well as a high level of performance.

When you talk about Dyn’s Internet Performance Management platform, what does that consist of?

At the core of everything Dyn does is we resolve DNS queries and we resolve them in the form of a managed service. Obviously, you’re familiar with the DNS query and the fact that when a DNS query is resolved you have an opportunity to make a number of choices about how you direct traffic, especially in complex network environments. We make those choices intelligently and we make them in a way that guarantees an optimized outcome – a performance optimization or a cost optimization. We do that with another component of the platform which is built around data.

+ Tech Q&As: The IDG Enterprise Interview Series +

Dyn has a unique set of data collectors that are distributed throughout the internet infrastructure out at the edge, in transit providers and elsewhere, where we are able to collect hundreds of billions of data points a day monitoring how traffic is flowing through the global internet. We use that information to help us make good decisions about how we steer traffic. That’s how we can optimize for performance or cost or routing, even geopolitical optimizations to avoid certain country borders.

There is a third broad capability in the platform which is around visualization and understanding of the network infrastructure as you’re deploying it, the reachability of it, the performance of it and the topology of your network infrastructure as you’re using perhaps multiple CDNs or multiple cloud providers. People really want to understand how traffic is flowing and they’re trying to optimize performance, geopolitically or cost-wise across these distributed assets. The three big tenets of the platform – DNS, data and data as a service and the network visibility and analytics – really tie together to create what we call the IPM Platform.

What do your Internet Intelligence products do?

Internet Intelligence (II) is the set of products that provide this level of transparency and visibility to people who are deploying network infrastructure, specifically outside-the-firewall network infrastructure. The Internet Intelligence products are a front end for many of the data offerings that we have. They allow you to see how traffic flows, they allow you to see the availability of various resources, they allow you to compare performance of different CDN assets, they allow you to compare the performance of different cloud assets.

The Internet Intelligence products are about creating that portal of visibility into assets, performance of assets and how assets are utilized. The information that is being displayed, the content that is being displayed through II, is largely the content that is collected by Dyn’s data services, though many customers also include some of their own data services.

How do you capitalize on all this intelligence you’re collecting from all the customers using this cloud-based offering? What can you share with people and what can’t you share out of that intelligence?

There’s a lot that you can leverage from the data. Think about the idea of evaluating CDN assets and making decisions based on performance or cost – which CDN is the most effective not just globally but by specific geography. I think that’s a very interesting one. If you want to leverage different distributions of various cloud providers, different geos of different cloud providers, we give you the visibility to make those decisions most effectively. Again, the benefit is derived in the form of cost and cost mitigation and performance and then resiliency. The value proposition always translates really into those fundamental areas.

Can you go into a little bit more detail about the traffic shaping products and what problems those solve or what opportunities those create for customers?

Most of the traffic management capability that Dyn provides is built around DNS-based traffic steering. We use DNS at that fundamental layer of the internet as a platform for traffic steering decisions and steering decisions are based on policy. Policy could be that we want to steer based on avoiding certain geographies, or we want to steer to resolve a particular shape of traffic we’re observing that is manifesting itself in the form of a DDoS attack. We would steer around that. Or we want to provide steering based on real-time performance data that we’re collecting through our data products. That’s, I think, a good example where you see the power of integrating the managed DNS platform with the real-time data collection and closing the loop using data to help apply the policy to achieve the objective.

I think you said that that kind of traffic shaping is done in real time based on policies. Is that correct?

That’s correct, yes. We use the data assets to inform the policy and we use the managed DNS platform to enforce the policy.

So, Jim, do you see that at some point becoming adaptive and learning and being able to do that more on its own as conditions change on the internet?

In fact, that’s what we’re doing right now. We think this is one of the very interesting things that tie together the value of the data asset and the value of the steering. Today, we literally collect, as I said, hundreds of billions of data points to profile performance on the internet. We use the results of that real-time data collection to apply a performance-driven policy in traffic steering. That loop is closed right now in the Dyn product line.

+ MORE ON NETWORK WORLD Slow Internet links got you down? It’s Dyn to the rescue +

I also wanted to ask about that multi-CDN capability. Can you talk a little bit more in depth about that and what that allows people to do?

What’s the objective in multi-CDN for a customer? The objective always is: I want to make sure my content is located in a way that it’s delivered in a high-performance manner to the end customer and to do that cost-effectively. With the II products we give the customer tremendous visibility into the performance of the CDN choices that they have and we allow them to play ‘what-if’. What if we use either a different instance of our current CDN or what if we use a different CDN provider altogether? One of the benefits of Dyn as a stand-alone monitoring business is that we create the opportunity for people to see across CDN providers as opposed to seeing inside of just one CDN provider. Typically, from the CDN providers you can get a lot of visibility about that CDN’s network and performance but it’s very hard to get a more neutral view across multiple CDN providers. We provide that capability so that our customers can look at CDNs and trade them off one another and make the decisions that optimize their costs.

Jim, I’m going to shift to some questions about the company and the competitive landscape, but is there anything else important about the products that you would want an IT executive to understand?

Our customer base has substantial interest in Dyn’s products as enablers in this migration to hybrid cloud. Enterprises are leveraging an increasing number of cloud services and they are quickly realizing they lack visibility into how those resources play together and how they control the performance and resiliency of their applications in infrastructure that they don’t directly control. They’re moving workloads out of the captive data center into cloud environments and it is critically important that they be able to manage those new applications and provide the level of service that they’ve been able to provide through their captive data centers. Helping companies move to the internet and operate as they were able to within their own proprietary data center environments is an important theme in the Dyn value proposition today.

If I understand correctly, you started with a focus on SMBs and consumers. Correct?

You do understand correctly. The company has a long and rich history in working at this layer of the internet. One of the company’s first products was a dynamic DNS product which was sold largely to consumers and prosumers. It’s still an important product for us. We have about 800,000 customers today that use a product called Remote Access that gives you the ability to access a device that sits behind a firewall and likely has a constantly changing IP address. We give you a stable way to access those devices and we sell that today for websites, we sell a service portal online. It turns out it’s actually a very interesting asset for Dyn as we expand our value proposition into the enterprise because every one of those end users is producing data. That becomes an important part of the data asset that we use to gauge the movement and performance of traffic on the internet in real time.

What prompted the shift to the enterprise IT market?

What prompted the shift is the recognition that this enterprise migration to cloud is such an important movement. We really are undergoing a platform shift. We’ve seen a number of them in our careers and we’re seeing a platform shift now to cloud computing and distributed computing as a service. It’s very real and it drives a tremendous amount of change in how enterprise IT is configured and functioning.

From Dyn’s perspective, we have 15 years of data history on how traffic moves on the internet. Some of the largest internet-native properties are customers of Dyn, companies like Twitter and Netflix. We have a unique asset in terms of our core intellectual property and our core competencies around data and traffic steering. It makes tremendous sense for Dyn to continue its growth trajectory into the enterprise and into enabling the enterprise migration to cloud and hybrid cloud.

Dyn against the world

There are a lot of companies that compete with you in one form or another, but who do you view as your primary competitors?

You’re right. There are a lot of companies that compete in one form or another but if you look at the integration of analytics and visibility through II and then managed DNS and then the data that enables it all, we stand [alone] in this space. When you look at the larger players who are doing load balancing, we see, for example, an F5 with a large install base. We see companies like Riverbed. We see other companies who are in one facet of the business but we don’t see a lot of players yet who have put all the pieces together to proactively control how traffic is managed on the internet through a data asset as well as a steering mechanism.

Is there a single company out there that’s more of a cloud-based management company, a next-generation company that you see as similar or more competitive?

Not yet. We’re starting to see some of the larger vendors look at this space. IPM is a natural extension of the existing categories of network performance management and application performance management. You’re starting to hear some discussion from some players who want to migrate outside the firewall into managing and controlling traffic over the internet. What we don’t see in the market right now is anyone who has put the pieces together in a way that gives you the data asset.

Again, we have this nice asset of 15 years and a lot of customers who are generating data, relationships with transit providers and others that give us traffic data, BGP data and others. We feel like we’re very well positioned competitively as the company that’s uniquely put the pieces together. But there are players out there who are doing managed DNS as a service. There are players out there who are doing some of the network visibility stuff as a service. We just feel that we still stand very uniquely as having put those three elements together.

As you mentioned, many companies are embracing a hybrid cloud model. In that environment, are you pitching your products as a replacement for traditional management tools or an augmentation of what they already have in place?

Largely an augmentation. Hybrid cloud is probably the number one enterprise use case we see today. Customers have tools in place to manage infrastructure in their data centers and, in many cases, we have relationships where our data is sold as a service to augment the capabilities of their console and monitoring tools. Then, outside the firewall, we’re an augmentation strategy that allows them to understand the next level of how their applications are being deployed. It’s a cooperative relationship with many of the existing providers because the data we have allows them to have more visibility even from inside the firewall.

You mentioned this a moment ago but a number of these folks, traditional management vendors, are trying to make the leap beyond the firewall. Do you think they’ll be successful at it? What holds them back?

IPM is going to be a very big market opportunity and we’re starting to see other vendors begin to position capabilities in the space. The interesting question is what are the barriers to entry that providers will have to overcome to get into this space? Dyn’s capability with DNS and traffic steering through DNS is tried and true, it’s proven. We’re the number one managed DNS provider in the world. We have many, many customers that are doing it, so there’s a very strong, resilient infrastructure that is resistant to the DDoS attacks that people are experiencing, that is able to survive through those kinds of events as well as able to steer and control traffic based on policy. That’s hard to build. That’s one asset.

The other asset, the data collection capability, is also hard to reproduce. There are about a million customers of Dyn, all of which help feed our data assets. Those coupled with sets of relationships of data collectors in strategically located places around the world again, I think, create a significant barrier to entry into this space. Will they be successful? The large vendors who want to play in a market find ways to be successful, and so I’m sure over the next couple of years we’re going to see a very interesting set of market dynamics around internet performance management. You’re going to see the larger players taking a very active interest.

Analytics and the Business Side

You had great success in your previous role as CEO of Netezza. What did you learn from the data analytics world that’s shaping the work you’re doing at Dyn?

One of the main things you learn from the data analytics world is that analytics absolutely need to be applied to a very tangible value proposition. Sometimes, people get very excited about big data and big data analytics and all the interesting things we can do with machine learning and technology. At the end of the day, the most important thing is how that is applied. That is where value is created for the end customer.

We’re approaching customers from the standpoint of understanding their business and using the analytics, data collection and steering capabilities to drive a tangible value proposition which manifests either in the form of increased revenue or reduced cost. One of the big learnings from living in the analytics world is that figuring out how to monetize the value of, let’s say, performance is critically important. That’s something we did very well at Netezza, something we did very well at Endeca. It really forces you to figure out how your technology intersects with the business workflows and processes and creates value in them.

We’ve talked to lots of management tool vendors over the years and one of the things they always promise is that they’ll be useful both to the IT side and the business side. Can you see the intelligence and the information that’s being created by your tools would actually be used by line of business executives? Is that a goal?

I see more and more in the enterprise that the “IT side” and “business side” lines are blurring. We see that in the DevOps environment. The DevOps guys are the new power brokers in the enterprise. The traditional IT data center stuff is obviously mission-critical but when you look at where applications are being deployed in enterprises, they are being stood up and deployed by DevOps people who, in many cases, are operating in the line of business as opposed to in centralized IT functions.

Do I see the vice president of marketing at a large enterprise interacting directly with Dyn tools? Not necessarily. Do I see the marketing function or the selling function driving the acquisition of capabilities that can be enabled by Dyn products? Absolutely yes, I see it already. The interesting thing about that from the vendors’ perspective is it really does cement the direct linkage between creating business value in the form of reduced cost or increased revenue and the deployment of a technology. You can look at tools like Dyn and that’s pretty low on the stack.

DNS is something everybody takes for granted. In reality, we’re able to show how you can monetize the value of performance and resiliency and that gives us a very important lever to speak to the line of business. Like I said, I don’t necessarily imagine the line of business user interacting with the Dyn console but I absolutely imagine and see today the line of business user driving the acquisition of Dyn tools.

The other thing we know about management tools is that in a great many cases, they’re either not used or they’re not fully used because they’re difficult or they don’t do what they promised. What have you done from the user interface, user experience perspective to actually make sure that these are tools that people can use on a day-to-day basis and derive real value?

It’s a question I love because I think there are a couple of different issues we need to talk about as well as some specific things related to Dyn. More and more people are suffering from tool overload. You say people don’t use the tools because they don’t deliver on what they say. I think there’s some truth to that. I also, however, think that tools don’t get fully leveraged because they have too many of them. We just have too many management consoles for controlling our app environments. We have too many consoles for understanding how our network is performing.

One of the very important trends here is the idea of a more integrated view of the overall application performance environment inside the firewall as well as outside the firewall. One of the things Dyn has focused on is making tools available as services and making data available as services so that other tools, be they custom home-built tools or be they other commercial providers, can leverage the Dyn capabilities. Making sure that it’s available as a service within existing infrastructure.

The other important element is how do you deal with the user experience of the tool set itself? We are really focusing on the user experience in the context of the value proposition. How do I evaluate, understand, procure, deploy and provision internet infrastructure? We’re building UX and workflows that tie together data, bringing in data at the right point in the workflow so that the person can have the context that’s required to make the decision that they’re trying to make. We’re building it in the context of an integrated platform that ties together data, analytics, DNS, traffic steering policy and so on in a common user experience so that you’re not jumping from area to area. That’s a major thrust for Dyn right now, to tie that user experience together in a cohesive manner as well as making all of those tools available as services to other application environments.

The capabilities you offer seem like they would be great tools for cloud and service providers to be able to offer to their customers. Do you have partnerships in place like that?

We do. I can’t name them at this time but I can tell you that those are an important part of our strategy. For example, providing a service, DNS as a service through large provisioners is something that we are doing today. We have tens of millions of domains that are provisioned using Dyn DNS as a service not directly through Dyn, but through partners. That’s an important part of our strategy, to partner with both cloud and CDN providers to provide our capabilities as a service as well as what we do independently.

Is the goal to do that behind the scenes or are some of these going to be named announcements? Is this something that we’ll hear more about in the future?

Most of them are actually done behind the scenes. A lot of it is white labeling where DNS is provided by Dyn but it’s a service that’s provided by our partner. We probably won’t make a tremendous amount of fanfare about that.

What’s ahead for Dyn? What should customers expect over, say, the course of the next year?

We’re very actively pursuing the integration strategies I talked about, around the integrated user experience. You’ll see us continuing to enhance our data assets. We’re always looking for ways to collect more data and do more interesting things with it. You’ll also see Dyn really focusing on this area of creating the direct-to-line-of-business value proposition through the deployment and monetization of performance and resiliency capabilities that we bring to market.