Cisco Blogs


Cisco Blog > Perspectives

Upcoming Technology Trends at Cisco Live

Cisco Live in San Diego is right around the corner. It’s the place to be to meet with people, learn and to stay current with the technology trends of the industry. What are some of the upcoming technology trends to watch out for at Cisco Live.

Software-Defined WAN (SD-WAN)

There is a lot of buzz about Software Defined Networks (SDN), Software Defined Data Centers (SDDC) and everything you can possibly think of and then adding software defined in front of it. Many of these technologies are not mature yet but SD-WAN is a viable technology as of now.

Cisco is realizing the SD-WAN through its technology called IWAN. IWAN is used when connecting to multiple Service Providers (SPs) and can more effectively work in such a setup than with vanilla routing. IWAN can choose the best exit, based on metrics such as latency, jitter and packet loss, which is not feasible with normal routing. It does this through a technology called Performance Routing (PfR). This technology was very complex in the past but has evolved to a much simpler configuration in its current revision. It can also help organizations save money by running DMVPN over the Internet instead of buying more costly MPLS circuits from the SP.

Provider Backbone Bridges Ethernet VPN (PBB-EVPN)

PBB-EVPN is mainly a technology for SPs or for enterprise that is running their own MPLS network. Building scalable multipoint layer two networks is always a challenge and has often been realized through Virtual Private LAN Service (VPLS) in the past. VPLS suffers from some drawbacks like explosion of MAC addresses, the requirement of a full mesh, the handling of multicast traffic and so on.

PBB-EVPN addresses these drawbacks of VPLS by using BGP as the control plane protocol, allowing for arbitrary topologies, implementing BGP policies for traffic engineering and the well-known stability and scalability of BGP. It is also designed to handle multi homed layer two segments which has been a challenge in traditional deployments. EVPN is also getting consideration to be used as a Data Center Interconnect (DCI) protocol to build scalable data centers.

Software Defined Data Centers (SDDC)

Data centers is one of the first Places In the Network (PIN) that is moving to a more software defined forwarding paradigm. The reason for this is that traffic patterns are fairly easy to predict where traffic is more of east-west nature compared to north-south in a normal campus area. The amount of traffic is massive and there are not many different types of devices that need to connect to the network compared to the campus.

Cisco’s solution in this space is the Application Centric Infrastructure (ACI) which is a software and hardware based solution available on the Nexus 9000 platform. With ACI it’s possible to define policies, which tiers can communicate, should the traffic be load balanced, how is traffic to the outside handled and a lot more. This is then programmed to the network devices that are normally in a leaf and spine topology by the Application Policy Infrastructure Controller (APIC). There are already a few DC’s running ACI technology and expect more news on this front at Cisco Live as the technology becomes more mature.

Segment Routing

Multi Protocol Label Switching (MPLS) is the de facto standard used by almost all SPs for forwarding of traffic. Normally labels to reach the PE next-hops is assigned by the Label Distribution Protocol (LDP) or Resource Reservation Protocol (RSVP). LDP is most commonly implemented but running LDP is an extra protocol that is simply assigning labels. Couldn’t this be handled by your IGP assigning the labels instead? Yes, and that’s the main idea of SR, to cut down on the number of protocols in the backbone and to allow for traffic engineering that is commonly only implemented through the use of RSVP-TE which is a complex protocol and that has scalability issues when deployed at large scale. SR is trying to solve some of these issues and software has been released to support this feature, expect it to gain more traction in the field as the software gets more mature.

Evolution of Enterprise Networks

Some people may argue that very little is happening in the enterprise networking space, which may be true to a certain extent but there are also technology trends in the enterprise as well. The main trend is to minimize the impact of layer two by building networks based on technologies such as Virtual Switching System (VSS), Virtual Port Channel (VPC), stacking and so on.

Cisco has also introduced the concept of Instant Access (IA) which is a similar technology as the Fabric Extender (FEX) available on the Nexus platform. With IA it’s possible to have access layer switches connected to the distribution and with the access layer devices acting as remote line cards. This creates fewer points to manage, gets rid of STP in the access layer and allows for technologies such as MPLS to extended to the access layer.

Network Function Virtualization (NFV)

NFV is another very hot topic right now. Routers and switches have almost always been physical devices but now we are starting to see virtual devices such as the CSR1000v, ASAv, Nexus1k, vWLC and many more. Virtual devices are a very good fit in some cases such as a Virtual Route Reflector (vRR) because it is easy to throw memory and CPU into a server compared to buying a router which may have less horse power. As this device is not in the forwarding path, all it needs is to have a powerful control plane and a device such as CSR1000v is a very good fit in this use case.

There is also an upcoming virtual IOS-XR device called XR9000v. There is already another XR platform available which is called XRv but the new XR9000v has much more of a forwarding plane and can achieve very respectable traffic levels. The XR9000v can then be deployed in samller Points of Presence (POPs) or in places in the network where it fills a specific role, such as providing a certain service to the network.

There are a lot of announcements coming up at Cisco Live both regarding new products and new technologies/features. If you can’t make it to San Diego, stay aware of new trends on Twitter, Cisco blogs and of course via the Cisco Live portal which will live stream some of the events. I look forward to meeting readers of the blog at Cisco Live. Don’t be afraid to say hi!

Tags: , , , , , , , , ,

Where to learn ACI

I remember walking through Cisco Live last year in San Francisco and hearing all about ACI and the Cisco DevNet program. To be totally honest, I shrugged it off as just something that was trying to get hyped up and would not have any real impact on what I do.

Well…that has changed over the last year for me. What made it change? I guess a desire to learn and grow. I have also seen how learning to code is becoming very important to any IT position. There is a lot of power in the software layer and learning to harness that power is very important to be able to accomplish our jobs. So, with that in mind I have started to learn Python. Why python? I come from a scripting language background with my years as a Windows admin. I never jumped in with both feet, but I did enough to be dangerous. And learning to write code has always really interested me, I just didn’t have a good reason to do it. Now, with ACI and SDN showing promise and no longer buzz words to me I am going to dive in and learn. Read More »

Tags: , , , , ,

Importance of naming standards in Cisco Unified Communications Manager

In network engineering I have learned that the biggest lie I tell myself is that “I do not need to write this down.” That being said, when you are in the heat of troubleshooting a production issue I really try to design my systems so that I can tell what the heck something does by a label or good name. This does not replace the need for other documentation, but it does help when you are in the heat of troubleshooting a system problem. As I started supporting Unified Communications applications, I discovered there are lots of opportunities to really create a mess when you are configuring things if you do not keep supportability in mind. I want to share with you some tips that I have found helpful in naming objects specifically in Cisco Unified Communications Manager; however, similar concepts can be used for other network components such as Access Control Lists on traditional network equipment too.

When you are starting with a fresh Cisco Unified Communications Manager install, you have a blank slate. This is both good and bad. Good in that you have a lot of flexibility in the system to configure things, but bad because if you don’t put some thought into naming it can get confusing quickly. Spending some time up-front will save you some headaches down the road. Even if you don’t have a fresh Cisco Unified Communications Manager installation, you can start cleaning things up as you provision new services and go back and adapt what is in the other systems when you have time to do so.

Some of the common things you will configure in Cisco Unified Communications Manager will be: Partitions, Calling Search Spaces, Route Groups, Route Lists, Route Patterns, SIP Trunks, Device Pools, etc. First let’s get started with some basic definitions of what some of common objects are. I will also share some examples of how I like to name things to keep them easily sorted so objects of similar function are grouped together in a long list. These are just examples, and your naming convention will have to be something that works for you, your team and your specific environment. Read More »

Tags: , ,

A Tour of Cisco’s Allen Data Center

As a consultant I have seen many different ‘Data Centers’, from Co-location facilities, to in house and well thought out, to a dirty closet that no one was using. Douglas Alger gave us a tour of Cisco’s Data Center in Allen, TX about a month ago. I was expecting to be impressed and I was not disappointed. Cisco has made a commitment to all of their Data Centers at least Leeds Silver certified. The Data Center in Allen, TX is Leeds Gold certified. Also, Cisco tried to use as much off the shelf components as possible so that this model can be replicated to every Data Center.

Outside of the Data Center building

When driving up to the Data Center it was not the usual look of a Data Center. You really have to know where you are going to find it. The building is surrounded by berms 15-20 feet tall. This is doubles as a camouflage for the building, but it’s primary purpose is to deflect tornados from hitting the building directly. If a tornado is heading for the building, the base would have to climb the berms which in turn would cause the tornado to ‘jump’ over the building.

The roof of the Data Center has a high level of wind tolerance, but the building is constructed in several layers. A tornado could take off several of these layers and the Data Center could continue to operate.

There are the typical barriers expected in a secure facility such as fencing, vehicle barriers, cameras, and a bicycle rack. Yeah, a bicycle rack. Part of the Leeds certificate is the ability for alternate modes of transportation to the office. Installing a bicycle rack and shower inside was an easy way to get additional points for the Leeds certification. Read More »

Tags: , , ,

To flow or not to flow?

NetApp’s newest storage operating system, clustered Data ONTAP (cDOT), leverages a backend of Cisco Nexus switches for it’s cluster interconnect network.

When configuring the switch/cluster ports for use with cDOT, the best practice is to turn flow control off as per TR-4182. In fact, that happens to be the recommendation for normal data ports as well. Why is that? Before we get into that, let’s cover the basics…

What is flow control?

Flow control is a mechanism used to help manage the rate of data transfer between two devices. This is done to help prevent a source evice from overwhelming a destination device by sending more packets than the destination can handle. These scenarios can occur if a source device is faster than the destination device (CPU, RAM, NIC, etc). This can also happen if the source is intentionally trying to flood the destination via a malicious Denial of Service (DoS) attack.

Flow control can be enacted for send or receive packets, or both. It can be hardware or software based. It can occur at multiple layers of the OSI model

For a real world analogy to flow control, think of how dams work. A dam will be installed to control the flow of water on a river, usually to create lakes or reservoirs. Dams can be used to adjust the water flow to prevent flooding, depending on rainfall. Network flow control does pretty much the same thing – it prevents data floods. Read More »

Tags: , , , ,