- SD-WAN as a Service: Meeting Businesses at the Business Level
- As SD-WAN Enters Mainstream, Security Features Transform the WAN
- SD-WAN: When the Solution Is Greater Than The Sum Of Its Parts
- Moving To The Cloud – Network Nightmare or Dream?
- SD-WAN: Steering Apps In The Right Direction
- Rolling Out SD-WAN at REI
- Taking SD-WAN Even Wider at Acadia
- Treating Your Cloud Like an SD-WAN Branch
- Succeeding With SaaS and Viptela Cloud On-Ramp
- The Complex Simplicity of SD-WAN
- SD-WAN Changes the Internet Security Model
- Approaches to SD-WAN Managed Services
- SD-WAN Fabrics Aren’t Interoperable. Should Organizations Care?
- The Current State of SD-WAN in Service Provider Networks
What does your computer desktop look like right now? Is it filled with a lot of applications that are installed locally on your machine? Or is it starting to look more like a cloud service portfolio? With the rise of offerings from Microsoft, Adobe, and others, the modern desktop application suite is running more in the cloud that it is on a local machine. And that’s just the applications on a desktop.
The increasing number of applications that are being offered as a service to support online collaboration or rapid updates has grown substantially in the past few years. Office365, Adobe Creative Suite, and may others are now available online and require connectivity to use the features of the software to their fullest potential. Users depend on these new functions in order to get their work accomplished.
Gone are the days of being productive even when the Internet was down or connectivity was impacted. Networking professionals need to make sure that their Internet connectivity is up and running at all times as well as making sure that the connectivity to the cloud is running at peak performance. With Software-as-a-Service dominating the desktop, can networking professionals keep up?
One of the biggest issues that you’re going to face as you move your organization to a SaaS model is the lack of visibility you have into cloud services. If you thought your IT organization was a black box before, you’re going to be shocked at how opaque it is inside the public cloud. Sure, you may have an idea of where the data goes at the end of the trip, but the entire path of the trip is difficult to determine.
And it’s not just your path to the SaaS application you have to worry about. With a centralized server running applications you only have to worry about the path to your data center. But with SaaS running out on the public cloud or the public Internet, every path your branch users take to get to the SaaS application can be a problem. Your main headquarters location can have good connectivity to Microsoft while your branch locations are traversing a service provider router that has packet loss issues. On top of that, you won’t be able to troubleshoot from your location because the path to the application isn’t the same.
The problem is that network systems have never been optimized for application visibility. In the past, a packet was just a packet. So long as the packet was delivered at some point everything was perfectly fine. That worked for the majority of data being sent over the Internet. Then, as applications took hold and application priority became more of an issues we started worrying about the packet path and the latency for the packets to arrive at their destination.
While some applications like voice and video are extremely sensitive to latency, other applications are more reliable when network conditions are poor. However, even small issues with latency on these resilient applications can cause poor user experiences when the majority of the communications must take place over the Internet. With the lack of real QoS over the Internet, there has to be some kind of solution to ensure that applications go where they are supposed to go.
Getting Onto The Cloud Highway
The solution to many of these connectivity issues is a cloud on-ramp. It’s a term that has been popularized by Viptela, and during the recent FutureWAN conference there were a couple of presentations about using a cloud on-ramp to help ensure that SaaS applications are running at peak performance.
The key to a proper on-ramp solution is monitoring. It’s not enough for you to know that a circuit is up and operational. You need to know the health of that circuit. You need to have some kind of metric that allows you to determine how that circuit is performing against other circuits so you know how the users are being impacted in all areas of the network.
Viptela has a metric like this called vQoE, which stands for Quality of Experience. This score is assembled by the Viptela vEdge routers as they query the SaaS providers in real time and assemble up-to-date reachability metrics. This means that Viptela knows how each link is performing based on where it is located. This direct Internet access (DIA) circuits are the most important way for users to reach SaaS applications. Knowing their status and health instantly is key to keeping your organization running smoothly.
But an on-ramp solution shouldn’t just give you the capability to see what the circuits look like. It should also give you the ability to reroute traffic in the event of failed or degraded links. In the case of Viptela, the intelligence built into the on-ramp solution allows you to bypass the local DIA exit point and send traffic over the Viptela fabric to a new exit point. In some cases, this could reduce latency by a substantial amount because traffic is exiting closer to the SaaS provider. In other cases, this can help distribute load in the event that local ISPs are experiencing higher traffic patterns.
Viptela Cloud On-Ramp effectively turns your branch offices into multiple DIA points for all your branch locations. With the analytics and health scores being offered by the dashboard, you can quickly see which users are having a poor experience and move them to a better circuit with just a few clicks of the mouse. And you can do this on a per-application basis to ensure that a diverse set of SaaS providers are being used effectively.
Bringing It All Together
SaaS shouldn’t need to be rocket science. Moving applications to the cloud should be simple and make everyone more productive. Likewise, the networking required to make those changes shouldn’t be difficult either. The network should be smart enough to know what’s happening and give you the capability to use that intelligence to make good decisions about traffic routing. Helping the network professional make good decisions about SaaS routing keeps everyone happy and makes you look like a hero at the end of the day. And that’s all it takes to make SaaS successful in your organization.