Videos: SD-WAN vs MPLS Network Technology Explained

In this post we take a look at two different network technologies; SD-WAN vs MPLS. Both are still widely used in networking today. There’s a lot of different opinions out there on which is better.  A Software Defined Wide Area Network (SD-WAN for short), is certainly the latest of the two services. Whereas the history of MPLS (Multi Protocol Level Switching) dates back to the early 1990’s.

  • Which has the better performance for your network?
  • What are the connectivity requirements of each?
  • Which has better security?

All good questions. We have two explainer videos below to answer those questions and more. Let’s take a look. Here’s an explainer video from Pat Herron at Chicago based TMSP Nitel. In the below video, Pat takes an objective looks at MPLS vs SD-WAN, as network solutions.

Video transcript: Thanks for joining this segment of From the Guys in Orange. I’m Pat Herron, the Vice President of Product Management here at Nitel. From the Guys in Orange is a video series we’re putting together to help our partners and their customers learn a little bit more about the topics of networking.

One of the questions we get asked quite a bit is, “Will SD-WAN replace MPLS as the leading network technology?” A lot of agenda and a lot of voices and a lot of opinions on this topic. But we’re going to look into some facts. What makes an MPLS network or an SD-WAN network? What are the features that make one more compelling than the other?

That’s how we’re going to spend our time today. To provide a framework for this discussion, let’s assume that the MPLS network is being provided on one circuit by a service provider over a carrier-grade private network with a typical service level agreement. It’s also assumed that the SD-WAN is being provided over two internet circuits. One, a carrier-grade dedicated internet service and the other cable provider broadband internet service.

SD-WAN vs MPLS: Which Has Better Packet Delivery?

In order for any application to work, the data packets that make up that application have to get to their destination. A carrier-grade MPLS network will offer packet delivery in the 99.9 percent range, meaning that only one in a thousand packets is lost in transit. This is typically backed up with a Service Level Agreement that puts money behind that performance metric. Data delivered over the public internet typically can have one percent or more packet loss, meaning that over one in a hundred packets may be lost along the way.

A few broadband providers offer meaningful Service Level Agreements for packet loss and during times of heavy network congestion, they may even have higher levels of packet loss, which makes having two or more internet connections super important for an SD-WAN solution.

The sd-wan solution, depending on the technology and the service provider delivering it like Nitel, can use methods to mitigate packet loss over the SD-WAN. Methods like Forward Error Correction that inserts parity bits into packets and those parity packets into the data flow allow for the SD-WAN to recover packets that may be lost along the way. So net-net. An SD-WAN solution, if designed well and implemented correctly, can offer packet delivery on par or at least very close to what an MPLS network can provide.

Application Prioritization and Performance: MPLS vs SD-WAN

Let’s turn our attention to application prioritization and performance. One of the benefits of having a private data network is that you get to prioritize which applications are going to flow over the shared resources and in that way, you can assure that applications will perform well. An MPLS network can offer three or four or even seven queues. These are priority tiers into which applications are mapped. The challenge for the network administrator is to identify and tag those applications and then map them into the appropriate queues and approach what we call “quality of service” or QOS.

This tagging and mapping can be both challenging and cumbersome. But it’s a proven approach that IT professionals know and rely upon. On an MPLS network, the quality of service is honored from end to end on the network so that the prioritization that goes in on one end is honored throughout the network and then still honored at the far end. This is very reliable and IT pros really like that level of reliability. An SD-WAN solution may offer dozens of queues into which applications can be prioritized and SD-WANs come with tools that allow for the identification of hundreds, even thousands of applications which really simplifies the mapping of those applications into queues.

This makes the task of prioritizing applications for the SD-WAN network a lot simpler both to set up and to make changes. Plus an SD-WAN solution with a good analytics and reporting platform can provide the IT pro with clear visibility at each application’s performance. As good as this sounds, when an SD-WAN prioritizes applications to be carried over the internet, they’re prioritized when they’re leaving on one end. Then they’re subject to the worldwide wild web which may not mean those packets are still in the same order when they arrive at the far end.

So what’s the net-net? Application, prioritization and performance is more guaranteed on an MPLS network than they are on an SD-WAN solution. But an SD-WAN solution provides a lot more tools and can make the act of prioritization a lot simpler.

MPLS example service prioritization by fastmetrics

Example MPLS application and service prioritization by Fastmetrics

SD-WAN vs MPLS Reliability

Let’s turn to reliability. There’s a short story on this one. Two circuits are going to be more reliable than one. Many businesses back up their MPLS networks using a failover solution to an internet connection using DGP and a VPN to then regain access to that private network.

The challenge with this approach is that businesses are paying for backups that they don’t really have much of a use. They’re not always tested and may not work as expected when the time comes. Even if they do, it takes a little while to fail over. So session-based applications like phone calls could be interrupted. The SD-WAN solution relies on multiple internet access links and is fully aware of the condition of those access links at any given point in time. Not just hard-down conditions but also degradation of circuit quality.

High priority applications can be routed over the best available path at any moment, including the moment one of the access links fails. Depending on the SD-WAN solution design, like the ones from Nitel, even a phone call will not be interrupted. So net-net, the SD-WAN solution provides a simple and powerful way to improve application reliability.

Is SD-WAN or MPLS More Secure?

Let’s turn our attention to security. MPLS networks are considered secure because only MPLS nodes that are reading the packet labels destined for that node can see the contents of the packets and the MPLS network is provided by one trusted service provider. So even though it uses shared network resources, businesses have relied on MPLS to provide secure site-to-site communication for years. SD-WAN creates a secure tunnel for packets to traverse between sites of the network using a form, a virtual private networking connection, such as IP sec.

Since there are multiple paths between the sites, traffic can be distributed across the paths, making snooping even more difficult. So even though it’s riding over the public internet, the SD-WAN provides proven methods to keep the application traffic secure and depending on the SD-WAN provider, additional network security services can be layered on top of the SD-WAN service.

For example here at Nitel, we offer a next generation firewall and a unified threat management service on the same platform, the same hardware and the same portal as our SD-WAN service. This really adds to simplicity. It can improve cost performance. So net-net, is SD-WAN or MPLS more secure? Both employ proven methods to keep customer and business data secure.

Will SD-WAN Replace MPLS?

So will MPLS be replaced by SD-WAN and if so, when? A recent market survey done by a leading global cloud network provider found that 29% of IT leaders indicate that their companies have deployed or considering deploying SD-WAN in the next 12 months. Plus 30% more are considering SD-WAN but don’t have a date in mind. But what’s really interesting is that of those, 62% report MPLS investments will increase or remain unchanged over the next 12 to 24 months and 53% expect their SD-WAN deployments are going to drive increases in their investment in network security.

So what’s the upshot? The move towards software-defined networking is undeniable. The growth rate and adoption over the coming years will be accelerating. But that doesn’t mean that MPLS is going away anytime soon. It’s a tried and true technology that’s relied upon the world over by IT professionals and it’s going to be around for years to come.

Hopefully this video provided you with a little bit of information about how software-defined networking is shaping our marketplace and I hope that you check back frequently to Nitel’s YouTube Channel to find out more from The Guys in Orange.

What Is SD-WAN & Why Do You Need It?

In this next video, Drew Schulke, the VP of Product Management at Dell EMC’s Networking Business Unit, looks at how networks have evolved. Drew also explains why SD-WAN technology forms a core component of Dell EMC’s hybrid and multi-cloud strategy.

Video transcript: Hi everybody. My name is Drew Schulke. I’m the Vice President of Product Management at Dell EMC’s Networking Business Unit and we’re here today to do a Lightboard session on the topic of why software-defined WAN or SD-WAN for short. So the goal here today is to really frame up your knowledge and understand why SD-WAN is such an important conversation point for our customers today and why it’s an integral component of our broader multi-cloud and hybrid cloud strategy at Dell EMC.

So to begin this journey, we’re actually going to step back in time a little bit and think about how wide area networks have evolved over the past couple of decades. So let’s begin with an example where we have a remote office for a large company or organization and we’re in the year 2003. So I have my happy employee there and we need to think about what that employee is doing on a day to day basis in terms of the type of applications he or she is using.

So back in 2003, you probably were spending a lot of time in solutions like email. You probably had some sort of ERP system. Either you’re doing financials or HR and so forth within a single ERP system and then you probably had some custom applications that were unique to your firm. Now it’s interesting as we think about in 2003, each one of those applications in terms of architecturally where they actually ran on infrastructure, they were all heavily-dependent upon infrastructure that sat inside some kind of centralized data center.

So in the case of email, you probably had a Microsoft exchange server. In the case of ERP, that might have been a mainframe or inclined server, maybe another dedicated server for that and your custom apps probably ran on a bunch of bare metal servers in that data center. So what that meant is for my employee in that remote office to be effective and productive, I needed a very solid connection between that remote office and that data center. By the way, that’s not the only remote office in this firm. I have others out there as well and they also need very reliable connections into this data center.

The way this was accomplished back in 2003 was I would typically use what was called a dedicated MPLS circuit. What you need to know about MPLS circuits is the following. One, they’re very reliable. You purchase one of these from a service provider. There’s a certain level of uptime that you’re actually contractually guaranteed to get. They have great quality. So I could do things like voice and video over them without worrying about any degradation in terms of the quality of those as deliverables.

But they’re also relatively expensive in terms of the dollar per bandwidth. But because it’s so reliable and because the quality is so high and because I just need a few of these to get to my centralized data center, it’s still a very good approach. So that’s one component of the Wide Area Network back in 2003. The other piece was that in each one of these remote offices, I would have a dedicated piece of hardware as well as one within my data center called a router. So what you knew about routers is they tend to be based upon proprietary technology, very, very expensive and you manage them on a box by box basis.

So this is kind of how I set up my wide area network to connect my own offices but not everything ran in the data center here. Sometimes I needed to pass traffic outside the data center and so I would bring everything through the data center because data center was a way for me to consolidate things like security. So I could use it as my single point of entry into my corporation and then that includes access into anything out there that we might just call the cloud today. But back then in 2003, we probably referred to it as the internet.

This is what a Wide Area Network looked like in 2003. Now let’s fast forward to 2019. I’ve got a similar employee now in this remote office and let’s think about the type of applications that he or she is using on a daily basis now. Well, I’m still using email. I’m still using ERP. I’m still using custom apps. But I’m also probably spending a lot more time on things like social media and a number of other – I will just call it generally SaaS-based applications that have become an integral part of my day to day job.

If we think about how email and ERP changed, here’s what I mean. Well, email in 2003 was running on a dedicated server in my data center. Now I might be using something like Office 365 which is a cloud-based email solution. For ERP, maybe ERP is still in my data center or maybe I’ve taken components of my ERP like my HR system and I’ve shifted it to something like Workday or Again, cloud-based or SaaS-based applications that are primarily running from the cloud. If I think about my custom apps, I might still have a server farm in my data center where a lot of these run on virtual machines but a great many of them also may be sitting on things like AWS or Azure.

So a lot has changed in terms of the infrastructure that’s running my day to day business in 2019 from 2003. But the Wide Area Network hasn’t fundamentally evolved. I’m still passing all of my traffic over these dedicated circuits, which are very, very reliable but also very costly into my data center and then going out into the cloud to access all of them in terms of where they live and breathe on a daily basis.

So this presents an interesting problem because these applications that are based out here also drive just huge amounts of traffic. So these circuits that I purchased in the past are being tasked in terms of the amount of bandwidth and data that I’m passing them over on a day to day basis and it’s only getting worse each and every day. As well it’s still complex to manage this WAN as I have to go device by device by device. So I’m looking for some operational simplicity here. This is where software-defined WAN comes into the story and why software-defined WAN is the customer conversation that almost every one of our customers wants to have right now.

SD-WAN Provides Dell EMC With Central Control

So to begin with, software-defined WAN allows me to have central control. What I mean by that and specific to how Dell EMC is approaching software-defined WAN is we have a virtual cloud orchestrator which allows us to look through the end to end in the wide area network and understand all the end points. It gets to access all the end points by accessing x86 devices that have now replaced these routers at all of these remote locations as well as in the data center. A standard-based hardware piece of infrastructure that we in Dell EMC know very, very well. Very cost-efficient as well and what software-defined WAN does is through this centralized orchestration, it gives me access to all these end points, giving me end-to-end visibility of the entire wide area network.

In addition, there are also virtual cloud gateways out at each one of these major public cloud providers, SaaS applications and so forth and to be specific, a couple of thousand of those throughout the global network that we know here at this point in time on planet Earth. What we can do with that is start to get an end-to-end view of the network and think about more efficient ways that we can start to pass traffic as opposed to these very expensive and dedicated MPLS circuits.

One way to go through that is to take advantage of broadband. This is the same broadband that you’re probably familiar with in terms of what’s running in your house today, in terms of how you’re actually getting internet service into your house. Contrasting it with MPLS, reliability is lower, quality is lower. But the cost per bandwidth is also significantly lower, orders of magnitude lower. So what I can start to do through this centralized place where I can set policy in my Wide Area Network is think about the priority of my applications.

What do I want to continue to run on the dedicated MPLS circuits because it’s important that I maintain a level of reliability and quality and what can I start to shift to run over much more cost-effective broadband solutions?

This is compelling because the economics between those two are pretty significant. So if we were to just do by an example here, to give you a sense, and I won’t use absolute numbers because it varies by region and so forth. But if I’m running a 100 percent MPLS circuit today and let’s say this is my cost level right here, if I can move to a model where I can get to 50 percent MPLS and 50 percent broadband, I can cut my cost to about 50 percent of what it was over here.

Now this still allows me to maintain certain critical apps running over that MPLS circuit. But I’m doing it just on a smaller pipe with less bandwidth required. Eventually, the goal for a great many of our customers is to get to 100% broadband. If you can accomplish that, you’re getting down to about 25% of the transport cost that you’re spending today. Now these are real expenses for our customers. They’re driving a lot of their operating costs and this is one of the key reasons why it’s so compelling and so many of our customers want to have this conversation today.

So we kind of covered this, how do I get to this 50-50 mix. How do I get to this 100 percent broadband if the reliability and quality are so important on these dedicated MPLS circuits? Well, here’s why. Here’s where the great duty of software-defined WAN comes into play is that I can start to utilize multiple broadband connections and through the ability through a centralized orchestration and policy engine and deep visibility into the performance, the latency, the jitter and things like that are in each of the paths that I can direct my traffic over, I can start to choose the optimal path for all of my data on my Wide Area Network over broadband, guaranteeing that I have the level of quality and service that I’ve become used to and expecting of MPLS circuits.

Now that’s a journey our customers begin. We as Dell IT are on this journey as well where we’re starting to implement this over time. But this is the end state that a great many customers are going to desire. But the great thing about SD-WAN is we can transition them into that end state by providing the ability to offload this and move the applications in a measured way, in a safe way without impacting any of the performance.

So I hope this comes together and makes sense in terms of answering that question of why SD-WAN. Very compelling economics. The fundamental nature of the applications and where they run in our day to day lives as companies and organizations have changed and we need to evolve our Wide Area Networks to accomplish that.

Software-defined WAN is absolutely the best way to go about that. As well I hope it came through, we didn’t talk about it explicitly, but when we think about the multi-cloud and hybrid cloud journey which will be on-prem, which will be off-prem and constantly evolving and shifting around, you need a very flexible and software-defined solution such as software-defined WAN to enable that. So between those two pieces, this is why SD-WAN. It’s a customer conversation that you absolutely should initiate. It’s very well-received and the timing is perfect. Hope you found this useful. Thank you for your time.