Breakout - Airheads Macau 2013 - Microsoft Lync, Unified Communications, Cloud Application Control, and Aruba Networks

Editor's Notes

1. These questions will be the basis whether to use heuristics or SDN API.
2. When it comes to performance, there is no match to Aruba’s ClientMatch technology. As you know, there are a variety of different client devices out there running on different operating system, different driver versions, even different capabilities like 802.11 a or b or g or n as well as 11ac. Not all these devices are created equal and just one poorly behaving client can bring down the performance of the whole network. The fundamental issue that on a Wi-Fi network the client device is in control. They make their own decisions on which AP to connect to, how long to stay connected to that AP and when to let go leading to the well known sticky client problem. Problem with this approach is that the Client devices have a narrow view of the network and are generally making decisions that may not be in the best interests of the overall network. ClientMatch fixes this by enabling the Wi-Fi infrastructure to make decisions on behalf of the client while keep a global network wide view in mind. If you are talking on your cell phone while driving down the road, you are probably going through several different cell towers. As you pass the towers, your active call and your devices is being actively steered by the cell company to the best cell tower for your device. Similarly, ClientMatch enables the infrastructure to steer the devices to the best possible AP based on several different factors like device type, location of the device, signal to noise ratio in the vicinity of the device as well the load on the Access Point. You can see this in action on the animated slide here where the iPad is being steered to another AP. With ClientMatch, the goal is to improve the quality of every single connection which effectively boosts overall network performance providing users with a superior user experience. As you see on this slide, Aruba has already been granted a patent on this technology making it unique and highly differentiated. Without ClientMatch, an 802.11ac network will operate no different than a 802.11n network and users will not experience much performance gains. In a real world test, we observed 98% of the devices significant improvement in their Signal to noise ratio when ClientMatch was enabled on the network.
3. Aruba’s 802.11ac capable AP220 series Access Points are the first on the market with true gigagit support. With total aggregate capacity of 1.75Gbps, which is well over 1 gigabit, Aruba’s APs are outfitted with dual GigE ports for link aggregation. In addition, the 220 series can be operated with standard 802.3af PoE which is supported by most access layer switches. This means customers can extend the life of their existing wired infrastructure while upgrading to the latest mobility solution without having to make capital investments for wired infrastructure upgrades. What’s most unique about the AP220 is the ability to exceed 802.11n speeds in the 2.4GHz band. Aruba’s 802.11ac AP supports up to 600Mbps in 2.4 band which is an increase of 150Mbps when used with Broadcom based devices like Apple macbooks, iphones and ipads as well as Samsung galaxy phones. The 220 series AP will be available both in controller based and controller less versions at a list price of $1295. Lets touch briefly on what we mean by purpose-built
4. http://www.clker.com/clipart-transparent-magnifying-glass.html
5. These questions will be the basis whether to use heuristics or SDN API.
6. These questions will be the basis whether to use heuristics or SDN API.
7. 30:24 – 32:44
8. These questions will be the basis whether to use heuristics or SDN API.
9. These questions will be the basis whether to use heuristics or SDN API.
10. QoS for Lync voice & video flows Aruba has been the leader in supporting Lync traffic and was the first certified to fully differentiate and support proper QoS in Microsoft Lync. Other vendors are just now catching up to where we were with Lync, but we have already moved on here. The key challenge with Lync is to correctly set and propagate through the network the correct quality of service (Qos) for Lync chat, voice, video, desktop sharing, and file transfers.  Since Lync traffic is encrypted it’s challenging to differentiate between these different flows, and so QoS is often set incorrectly by the network infrastructure resulting in a bad customer experience.  In our first generation of Lync services we developed heuristics to look at the Lync traffic flows and differentiate between latency-sensitive voice and other services. This worked well and allowed us to set the right QoS when Lync traffic flowed over the network.  If we received Lync packets that were mismarked, say from a switch, we could also retag the packet.
11. Traditional wireless architectures map a client to a particular SSID, then segregate different traffic types by SSID and map each to a particular VLAN. Separate VLANs for voice and data implies a security model, while use of separate SSIDs implies different QoS handling. This model does not scale very well nor does it support converged applications or devices. In contrast, Aruba Mobility Controllers have built-in application layer gateways (ALGs) – Aruba’s Mobile Edge Architecture integrates a stateful application-aware firewall in the WLAN infrastructure. This firewall provides the WLAN with capabilities to detect most common VoIP protocols including SIP, H.323, NoE, SCCP, Vocera to make bandwidth reservations and QoS prioritizations as required.Integrated stateful firewall Segregate, prioritize and police traffic by session, not device, SSID or VLANSSIDs can be used for other purposes (security, radio characteristics)
12. QoS for Lync voice & video flows Aruba has been the leader in supporting Lync traffic and was the first certified to fully differentiate and support proper QoS in Microsoft Lync. Other vendors are just now catching up to where we were with Lync, but we have already moved on here. The key challenge with Lync is to correctly set and propagate through the network the correct quality of service (Qos) for Lync chat, voice, video, desktop sharing, and file transfers.  Since Lync traffic is encrypted it’s challenging to differentiate between these different flows, and so QoS is often set incorrectly by the network infrastructure resulting in a bad customer experience.  In our first generation of Lync services we developed heuristics to look at the Lync traffic flows and differentiate between latency-sensitive voice and other services. This worked well and allowed us to set the right QoS when Lync traffic flowed over the network.  If we received Lync packets that were mismarked, say from a switch, we could also retag the packet.