The biggest buzzword in the network industry is Cloud: the majority of organizations have a strategy to use cloud-based services and applications, whether it be Public or Private clouds. Organizations have come a long way since ‘migrating to the cloud‘ discussions began . Take a look at this video recorded just a few years ago when cloud was still an enigma:
But Cloud has never been a new concept: IT professionals have been migrating applications to centralized datacenters for decades now…mainly to share recourses and save money by having less IT personnel supporting branch offices. Unfortunately, application performance as well as reliability and uptime requirements quickly became barriers to this centralization. And while headquarters is where you often have the most skilled IT professionals and reliable facilities, branch offices often have the most customer interaction, requiring the most performance and reliability.
Today, WAN connectivity is much more reliable with higher performance, and bandwidth costs are dropping. This enables companies to have bigger links and backup links (that are more closely to the primary link). Companies can once again start migrating to a centralized datacenter, a private cloud, or public cloud applications.
The customer premises router (CPE router) is located at each branch, but often it is only providing WAN connectivity –and not many necessary services for Cloud migration.
You’ll now have to rely on your WAN connectivity for any day-by-day tasks, and if for some reason you lose it, it becomes instable or slow, your productivity and customer satisfaction will drop. You need the mechanisms to guarantee that WAN connectivity is always up, the correct applications are using your bandwidth, and the applications have good behavior even when used over a WAN, providing predictable performance for any applications, as well as survivability and fault tolerance.
Obstacle 1: Application Visibility and Prioritization
You need to have a better visibility of which applications are using your WAN:, what the needs are for each application, and control mechanisms to prioritize mission critical traffic so that correct application have priority over not unimportant applications. With embedded functionalities like NBAR, NetFlow, and IPSLA, Cisco enables network administrators to get an in-depth view of WAN utilization, and mechanisms like quality of service (QoS) and hierarchical QoS (HQoS) enable administrators to control the environment prioritizing traffic as needed.
Now we are going a step forward, with Application Visibility and Control (AVC) and PA (Performance Agent). AVC supports over 1000 applications using NBAR2, enabling a granular and smarter application control. It also has an advanced graphical reporting interface and management with CiscoPrime Assurance Manager and Cisco Insight.
Cisco Performance Agent (PA) measures application response times at branch edge, extending visibility into remote sites. PA is embedded on the IOS, so there is no need to change the router or add/change any equipment.
Obstacle 2: Application Acceleration
Once you have WAN utilization visibility and apply the correct mechanisms to ensure proper prioritization, you have to focus on how we can deliver applications to branches with the same performance as we have in the datacenter. Applications were written for a LAN environment – not to a WAN – and WAN latency increases application response time and degrades user experience. If now we have to access applications on the cloud, we have to create mechanisms that accelerate applications, providing similar performance for branch users.
TCP/IP optimization, compression, and caching has become a must have. Cisco Wide Area Application Services (WAAS) is the family of application acceleration from Cisco, from the equipments that is seated on the datacenter, to appliances for branches, or even integrated on the CPE.
Cisco WAAS Express is the solution for application acceleration embedded on IOS and is enabled via licensing, avoiding the need of an extra box on the branches or replacement of the access router.
Obstacle 3: Reliability
Consolidated data center and cloud environments also increase reliability demands and need to adapt to change in the network conditions. Also, the backup WAN link is often underutilized, wasting money and increasing convergence time.
Performance Routing (PfR) is technology used to guarantee the maximum usage of WAN links, performing intelligent load balancing based on applications needs, and providing a much faster convergence time in the case of error or failures.
Using IPSLA, NetFlow or even RTP, Cisco Integrated Services Routers Generation 2 (ISR G2s) keep a table of link conditions, such aasjitter, delay and latency, and perform load balancing based on those conditions and on the application needs. This ensures that all WAN links are being utilized in a manner that applications uses the best link for their needs, while also ensuring a much faster convergence time if one of the links fails.
Obstacle 4: Survivability
If everything goes wrong, you’ll still need to have mechanisms to stay productive. You can extend the datacenter to the branch, extending virtualization to the remote office and guaranteeing survivability for applications. This is not only important for productivity, but is also a requirement to stay in compliance with several policies that still may require local presence.
The solution that Cisco presents for this need is the UCS Express on the ISR G2, the extension of the virtualization to the branch router. UCS Express is based on the SRE (Service Ready Engine) blade and can be used in any router with an SRE slot, which means ISR G2 2911 or bigger. The SRE blade is a full PC that can be inserted o the router; it has its own processor, memory and hard-drive, and is internally connected to the router by two GB interfaces. On this blade, you can run the VMWare’s vSphere Hypervisor, and on top of it you can run any OS (Like Windows, Unix,..) and have your own applications running inside the router. UCS Express enables extension of the datacenter rto the branch router, providing survivability to applications and meeting compliances policies that requires local presence.
In conclusion, migration to Cloud is something that will happen. The good news is that today we have enough WAN bandwidth and reliability that enable this transition, and the Router plays an important role on this. Mechanisms embedded on the router allow us to have visibility and control on what is going on over the WAN, accelerate application ensuring an adequate response time, intelligent load balance mechanism ensuring the best utilization of all WANs links and faster convergence time, and finally extending the virtualization to the branch router guaranteeing survivability and policy compliancy. Cisco ISR is the only router in the market that can provide all of those in one box, ensuring a seamless migration to Cloud. For more information, please visit www.cisco.com/go/isrg2.