eG News

The notion that application and desktop virtualization are no longer buzzwords is a clear sign that these technologies have matured and gained mainstream adoption. However, the end-user computing landscape is ever-changing.  We continue to see the platform providers - Citrix and VMware - compete with protocol improvements and optimizations with a view towards improving the user experience, meeting user demand for graphics intensive applications, and accounting for the increased adoption of hyper-converged systems for scaling deployments. Plus, we also see an increase in cloud-enabled desktop and application virtualization that simplifies management and enables dynamic workload changes to be handled through cloud bursting.

From a performance monitoring and management perspective, here are five key predictions for 2016:

1. Enterprises will look to incorporate performance monitoring and management technologies from the outset as they deploy new and upgraded application and desktop virtualization. Traditionally, monitoring tools are considered only once an infrastructure has performance issues. But at that point, it is often too late because there is no empirical data to indicate what is causing the issue. Even when the monitoring tool provides sufficient visibility into bottleneck areas, re-engineering the infrastructure is often cost-prohibitive or otherwise impractical. With years of experience, enterprises, system integrators and virtualization architects now realize that performance monitoring is like insurance - it is something you need from the beginning - as you start your application and desktop virtualization deployments.  In doing so, organizations will better see performance issues and abnormalities early in the life cycle and have a practical opportunity to take action to prevent the issues from becoming service impacting. 
2.  As application and desktop virtualization vendors explore hybrid models, the need for performance management will increase. Application and desktop virtualization vendors are seeking ways to make it easier for organizations to deploy their technologies, for example, the Citrix Workspace Cloud, where the control plane of the infrastructure resides on the cloud but the data plane resides on premise. Although this hybrid model makes it easier for enterprises to set up the system (in this example, Citrix managing the control plane), monitoring and management of performance will not be any easier. On the contrary, there will now be an additional layer of complexity when a problem happens - is the slowdown due to the network, the on-premise data center or applications, or could it be in the control plane managed by the cloud service provider? This additional layer will require more visibility to take proper advantage of the benefits of emerging hybrid models. 
3. Performance monitoring and management tools will have to be more specialized for application and desktop virtualization than ever before. A significant number of organizations deploying application and desktop virtualization have tried to use legacy monitoring tools to meet evolving performance management requirements. The challenge here is that most general-purpose monitoring tools are not sufficiently specialized for application- and desktop-virtualized infrastructures. Such tools will not be effective for identifying and helping resolve issues in advance. As the rate of change in end-user computing increases, to be effective, performance monitoring and management tools will need to adapt quickly to these changes. As one example, many organizations are beginning to adopt graphics processing units (GPUs) to deliver an improved user experience.  So in order to help maintain a great user experience, performance monitoring tools must now be able to monitor GPU utilization to identify times when utilization is high and provide actionable insights: Why the utilization is high and which virtual desktops or users are responsible. 
4. An increasing need for efficiency will drive organizations to seek a single integrated console from which they can monitor their application and desktop virtualized infrastructures as a service. Historically, organizations have relied on tools provided by the virtualization vendors to manage application and desktop virtualization. Increasingly, the focus of these vendors has been on making their platforms perform better, which means holistic monitoring and management has received less attention. As a result, organizations have needed to deploy multiple management tools and consoles for monitoring the different infrastructure tiers that are associated with the application/desktop virtualization service. This has resulted in slow, manual diagnosis and an increasing reliance on experts to perform manual analysis and troubleshooting. As the industry matures, faster problem diagnosis is becoming a requirement, rather than a nice-to-have. Increasing automation in the performance management space, by correlating metrics from every layer and every tier of the infrastructure will become mandatory for application and desktop virtualization success. 
5. Performance monitoring and management will not just be about troubleshooting. Enterprises will start to use performance monitoring to fine tune and optimize their application and desktop virtualization rollouts to improve ROI. In most cases, performance monitoring tools are being used to troubleshoot problems as they happen. At the same time, these tools collect a wealth of information about the infrastructure that provides interesting insights that can help optimize the infrastructure for maximum capacity and ROI.  For example, instead of simply adding more hardware, by analyzing performance across the server farm, organizations can identify the key resources that are bottlenecks and discover precisely what type / how much capacity needs to be added to enhance performance and utilization.