Thanks in large part to investor-driven demand for constant double-digit growth, the WAN optimization market has taken a substantial beating, and as a result the term “WAN optimization” has become something of a lament. Vendors looking to distance themselves from the term have latched onto the “software-defined” bandwagon, and the result is “SDN WAN” or “SD WAN.” However, the need for WAN optimization is not going away; in fact, it is only going to increase because the WAN is a critical component of cloud-based deployment models. Cloud is not a magic fix for all that ails IT. The laws of physics still apply to cloud deployments just like any other and distance matters. As Grace Hopper demonstrates in her simple visual demo of what a nanosecond represents, it takes a fixed amount of time for data to travel over great distances. Therefore, when you decide to take all of your Microsoft Office applications that once resided locally on the end-user desktop and move them to a cloud-based model such as Office365, the distance between the user and the application has increased dramatically. Have we learned nothing from trying to virtualize the desktop? Desktop virtualization projects only moved the workloads off the desktop to our own data centers. Cloud deployments move workloads off the desktop to new geographic locations outside the data center, and possibly over public Internet. What could possibly go wrong?
Reliability and Performance Will Make or Break Public/Hybrid Cloud
In order for public/hybrid cloud to be successful, we must optimize WAN connectivity in an intelligent and automated fashion that takes advantage of both private and public network connectivity. Public and hybrid clouds change the connectivity model, and that creates challenges to visibility, control, security, and reliability. EMA’s April 2014 report, “Managing Networks in the Age of Cloud, SDN, and Big Data: Network Management Megatrends 2014,” revealed that 75% of the companies who had deployed into the cloud reported difficulty with monitoring and troubleshooting their cloud deployment. Clearly the new challenges presented by cloud can limit the effectiveness of traditional methods of monitoring, control, and optimization, or even render them completely ineffective. EMA wants to better understand what steps are being taken to improve visibility and performance of cloud deployments. To that end, we are getting ready to launch a new EMA study that looks at how customers are tackling network monitoring and performance in their cloud deployment environments. We plan to look at what types of workloads are being deployed to the cloud and try to understand what are the most difficult network and application performance issues being experienced in cloud deployments. We will also look at what network-based techniques are being used to monitor performance across various cloud deployments and see how well they are working. Finally, we want to know what, if any, optimization technologies are being actively used in cloud deployments and which are getting the best results.
The cloud deployment model creates new performance, monitoring, and optimization challenges. The WAN clearly has a major role to play, and it provides new opportunities for both traditional and cloud-based optimization solutions. It will be interesting to see which solutions are getting the most traction and how monitoring and optimization need to grow and change to meet help enterprise IT departments be successful in the cloud.