Lync Migration Featured Article
November 11, 2016
Is Your Infrastructure Limiting Your Digital Transformation?
By Sean Applegate
Director of Technology Strategy and Advanced Solutions
The latest evidence that enterprises are rushing to embrace the cloud comes from McKinsey & Company, which declared last month that, “The cloud debate is over—businesses are now moving a material portion of IT workloads to cloud environments.” Organizations are experiencing a digital transformation, and I don’t use that term lightly. As I explained in my recent column on creating a “high velocity culture,” strategy, not technology, is driving digital transformation. IT needs to embrace a systems thinking approach to designing the cloud-first business, and determine which technologies to leverage. The clients I interact with are now moving internal business systems to the cloud at a more rapid pace, where previously most of their systems were focused on customer engagement. Next, let’s examine how this enables you to take full advantage of the benefits of cloud computing with the help of some real world examples.
As you’ve likely discovered, migrating systems and applications to the cloud is much more complex than flipping the figurative switch. Cloud computing may not be new, but that doesn’t make building and managing a complex hybrid IT infrastructure anything less than revolutionary for your organization. That’s why you need a comprehensive strategy to direct short- and long-term decisions. After all, even something as seemingly straightforward as migrating users from Microsoft Office to the cloud-based Office 365 can create new and unfamiliar challenges.
Gartner (News - Alert) this year published an excellent guide, “Network Design Best Practices for O365” that begins with an explanation of why so many organizations struggle to overcome network constraints, particularly with users scattered around the world in remote locations. A common challenge arises from deploying O365 in a single geographic location, then trying to extend access to employees worldwide. The result is a network beset by high latency and capacity bottlenecks. Existing network paths (MPLS to Data Center to Internet) are simply not ‘good enough’ for most cloud usage scenarios.
The necessary first-step to any cloud migration plan is to understand and optimize your end-to-end performance for the cloud. This requires analyzing user performance requirements such as business transaction times, as well as network capacity, latency, etc. Be on the lookout for high latency and performance bottlenecks at aggregation points such as the datacenter and Internet peering points that will cripple users’ productivity levels. These challenges are most acute for employees in Asia Pacific, Africa or the Middle East who have to come back to the U.S. to use applications.
For example, if we add 100ms of latency along our cloud journey by requiring employees to come to the data center, be processed at a security stack, and then go put the Internet to the cloud, it can often double the business transaction time when compared to an on-premises application, which has much lower latency. Below is a network model using Transaction Analyzer, to estimate how a three second SLA would be impacted if an FTP server was moved to the cloud. In the first scenario, ‘Client to DC,’ the FTP server is hosted in the data center and accessed over a 10Mbps MPLS WAN path. The 1MB transaction completes in 2.7 seconds. The ‘Client to Cloud’ scenario takes over twice as long, finishing at 5.9 seconds. If we add WAN optimization with an estimated 75 percent data reduction, the result is a one second transaction, saving users significant time. If we had used SD-WAN to go direct-to-internet from the branch, the time would have been similar to the 2.7 seconds in our original baseline due to similar degree of latency.
Consider the case of Schneider Electric, a nearly 200-year-old French firm that specializes in energy management. It employs more than 150,000 employees in over 130 countries. MPLS and local networks connect 1,000 remote sites to multiple regional data centers, where many critical applications have been consolidated. In addition, the company has adopted cloud-based solutions such as infrastructure as a service (IaaS), provided by Amazon Web Services (AWS), and software as a service (SaaS (News - Alert)) such as Salesforce.
To establish visibility over the entire hybrid IT infrastructure, Schneider Electric leverages network performance management tools to create Internet traffic reports that make capacity planning and traffic engineering easier, as well as controlling costs associated with SaaS and cloud services providers. The result is high-level visibility into traffic movement based on Border Gateway (News - Alert) Protocol (BGP) Autonomous Systems (AS) numbers to help streamline application performance in a company's vast, hybrid IT environment.
Typically your WAN is the key constraint due to limitations of the speed of light. You must design around this limitation, and that’s where SD-WAN technologies are proving to provide agile, secure and cost optimized architectures for hybrid enterprises. You can expect to realize a 5x reduction in cost when using the Internet for cloud/SaaS/Internet-based traffic, and MPLS for only critical internal apps (i.e., UC, Point of Sale, CRM/ERP, VDI, etc.). Use this flexible SD-WAN cost savings calculator to estimate your own benefits. SD-WAN enables the use of intelligent path selection, application-aware quality of service, cloud-ready AutoVPN, embedded firewalls and web security gateways and real-time monitoring to create a closed loop performance process.
When Australian start-up SimplePay decided to take its payment processing service global, the first priority was to replace its existing bespoke networking infrastructure with one that would deliver the security and reliability the company needed to scale its agile business. The systems worked fine in Australia, but it quickly became clear that they could not scale worldwide, managing them would be too difficult, and performance levels too unreliable to meet users’ needs.
SimplePay partnered with Amazon Web Services (News - Alert) (AWS) so IT can spin-up infrastructure in regional clouds wherever required, and manage that infrastructure through a single dashboard. Then it built cloud-ready branches using SD-WAN as the secure performance-centric fabric between all of the virtual clouds SimplePay runs in AWS and its employees across the world.
Integrating WAN Optimization makes SaaS and cloud business transactions 2-33x faster by overcoming the impact of latency. The key is to make sure your SD-WAN and WAN Optimization technologies integrate so you maintain application-awareness end-to-end so path selection/QoS/performance visibility/SLA tracking work end-to-end. Build closed-loop performance processes into your operations. This allows you to monitor and troubleshoot sites, users and applications end-to-end. This provides huge benefits to reducing MTTR, validating investments, and building shared KPIs with your LOB.
Schneider Electric and SimplePay demonstrate how to take advantage of all the benefits cloud computing has to offer, without remote employee productivity impact, by building a high-velocity culture that facilitates collaboration among the application, infrastructure and business teams. The goals are to optimize business performance and reduce operating costs which result in key advantages in a very competitive market, not upholding outdated policies or bureaucracies which put you in a position to be disrupted.
Edited by Alicia Young
Article comments powered by