Martin Finlayson, Head of Visual Communications, avsnet
Real-Time Media is the real crux of what is so important about any cloud service. In fact, it is possibly the most misunderstood aspect of modern communication for one simple fact – IP networks were never designed to carry real-time data.
Now don’t get me wrong, it is entirely possible to design and build an IP network that is perfectly up to the task. We have plenty of engineers here who can give you chapter and verse on how to accomplish that, I might even convince them to write a blog on the subject if there is enough interest?
The problem is that most networks evolve over time. Organisations seldom have the luxury of building a network from the ground up to cope with the amount of data, and differing types, that the modern network must support. Even organisations that have diligently created a rock-solid and predictable network, risk undoing all that effort and expenditure by implementing a poorly thought through UC Cloud strategy. Why?
Simply because all the tools and technologies that have been developed to aid real-time traffic on your network are not available on the internet!
I want to discuss how network architectures can have a dramatic effect on the quality and viability of any UC solution you try to run over it.
So, what exactly do I mean by real-time media?
Real-Time Media, as the name suggests, are the streams of data that carry the bidirectional audio and video signals between the clients involved in the call. It’s a lot harder to describe the scenarios in text so here are some simple diagrams to illustrate the concepts and challenges.
Figure 1 shows the default situation with most modern UC cloud services in a point to point call. Although all the call control traffic always goes via the cloud service, the critical media streams stay on your LAN/WAN. This will provide the best service level to your users and potentially allows you to leverage those real-time tools and technologies I was referring to previously.
Sadly, not all solutions are this smart and Figure 2 shows what can happen even in a point to point call if the cloud service does not support direct media. In this instance, all the media streams go out to the internet and come back in again through your perimeter firewall. Cloud VMR services (Virtual Meeting Rooms) that support registered H.323 devices have to use this model for instance.
Irrespective of whether the service supports direct media or not, Figure 3 shows what happens the moment a third participant is added to the call. Despite this being a simple drag and drop process for the user when they add another participant, what happens in the background is that the entire call is dropped and rebuilt in the cloud.
This means that now all the media from all the clients is now traversing your perimeter firewall and going out to the internet and back again.
Note: This is typically the scenario that passes the PoC (Proof of Concept) phase of any testing due to the small number of users involved. However, when the service goes live the amount of internet bound traffic starts to cause a real problem and the service quality often suffers to the point of being unusable in peak periods.
Figure 4 shows the advantage of cloud services when you have a workforce that are not all based in a single location. In this instance, the cloud offers the perfect place for all the users to dial into as it is the best common location for all involved. This diagram demonstrates the importance of considering the size, type and disposition of your organisation when evaluating whether a cloud service would be a good fit or not.
When it comes to Cloud, one size most definitely does not fit all.
Finally, Figure 5. This shows the model that a hybrid deployment gives you. Here all the media on a single site stays on a single site, only the traffic destined for off-LAN users needs to be routed out to the cloud service. This offers the best compromise between the flexibility of the service and the quality that you are able to deliver to your users.
I trust these simple diagrams have helped you understand where a cloud UC service might be a good fit and where it might be a liability.
If the service doesn’t deliver an excellent user experience, people won’t use it. This means you won’t see the business benefits video collaboration could deliver to your company and will waste your IT investment.
This also creates potential risk! If the company’s video collaboration service offers a poor user experience, employees might deem it easier to use a public service like Skype. Then you have unsecure shadow IT in your organisation and the business risk factors this exposes you to.
Video collaboration can generate significant business benefits, but only if it’s implemented correctly and the network is designed to effectively carry Real-Time Media – is yours?