By: Frank Ohlhorst dnu
Virtual desktop infrastructures are supposed to make things easier, especially for IT staffers, who can eschew many of their desktop management chores as virtual desktops replace traditional ones in the enterprise. However, enterprise network administrators are finding managing VDI solutions has become a time-consuming, less-than-ideal process. As VDI implementations grow, so does the time needed to manage the various bits and pieces required to make VDI work.
With VDI, the management chores have changed from managing simple desktop PCs to managing a complex infrastructure of virtual desktops. The complexity of VDI stems from the various pieces of technology that must all work in concert to provide an adequate end-user experience. Those technologies include servers, hypervisors, connection brokers, virtual hard drives and, of course, all of the associated hardware, ranging from Ethernet connectivity components down to endpoints. For example, a VDI solution may use a Microsoft Windows Server, a VMware hypervisor, a Leostream Connection Broker and a display protocol, such as RDP (Remote Desktop Protocol) or PCoIP (PC over IP).
eG Innovations is hoping to remove some of the pain of VDI management with its eG VDI Monitor, a management application that enables administrators to understand what is occurring on their virtual infrastructures. eG VDI Monitor works as an end-to-end management product that monitors the various pieces that make up a VDI platform. It is its cross-vendor support that makes the product compelling to those looking to rein in out-of-control VDI platforms. The product works with multiple hypervisors, including Microsoft’s Hyper-V, Citrix’s XenServer and VMware’s vSphere, as well as with such connection brokers as Leostream, Citrix’s Xen Desktop and VMware View.
eG Innovations succeeds at placing important information at an administrator’s fingertips, which otherwise would be hard to identify using the tools that are often bundled with the individual products. eG VDI Monitor can be a powerful ally for administrators looking to prove that a VDI implementation was successful or, better yet, how to make sure it is successful.
eG VDI Monitor brings some interesting capabilities to the table. First and foremost, the product is able to monitor and report on user activity on the virtual desktop, not just how the virtual machine is performing. Most VDI management products are unable to look inside the virtual desktop to record user activity, much less report on it. eG appears unique in this capability, and the ability to monitor user activity may make a significant difference when trying to resolve problems on a VDI platform-that is, is it something the user did, or did the software fail?
Those capabilities are the gravy of the product, as the meat and potatoes portion includes end-to-end monitoring of all components used for VDI, including servers, routers, switches, connection brokers, hypervisors, display protocols, bandwidth use, throughput, latency and pretty much any other VDI component one could think of. However, those capabilities come at a price, namely a very complex installation and setup process.
While monitoring and identifying problems is an important management capability, eG VDI Monitor attempts to go one step further by offering information proactively. The product produces warnings and alerts based upon conditions that indicate a problem is forthcoming, before the end user is impacted. For example, if the product notices increased latency or a higher than usual transmission error rate, it will warn the administrator of the potential problem.
Working with eG VDI Monitor
Working with eG VDI Monitor
eG VDI Monitor is part of the eG Enterprise Suite, a comprehensive set of integrated utilities that monitor a range of products and services in use by an enterprise. The eG Enterprise Suite, via plug-ins and optional add-ons, allows an administrator to build out a customized monitoring and support solution. eG VDI Monitor shares many common requirements and installation needs with other members of the suite.
I tested eG VDI Monitor on a customized setup in the lab. Hardware-wise, I started off with an IBM Series x3200 server with dual Xeon CPUs and a Dell PowerEdge 2900 with a pair of Xeon CPUs. Both systems were running Windows Server 2003 and were configured with 16GB of RAM. From the software side, I decided to try out eG VDI Monitor’s capabilities with a Citrix XenDesktop 4 and installed the appropriate software components, as well as vCenter 4 and Windows IIS (Internet Information Services) 6.
Installation of eG VDI Monitor can be a complicated and daunting process that involves several steps and decisions, so it is best to approach the installation with a plan and review all of the documentation and capabilities before diving in. That said, the installation process did not throw any real surprises at me or come grinding to a halt, but I did need to be familiar with several technologies, including Microsoft server OSes, IIS, Citrix XenDesktop, Active Directory and other technologies that are common to a complex network. eG Innovations offers support, as well as authorized reseller partners that can help with installations and configurations.
Once the management suite is installed, the product is able to do an auto-discovery of virtual machines on the network (or more correctly, on the monitored server). Administrators have several tools available to them to deploy agents, go agentless and build groups of monitored components. Once again, configuration takes a bit of effort and is best accomplished by someone with the appropriate technical knowledge.
After dealing with the complexities of the installation and configuration process, I was pleasantly surprised at how easy it was to use the management console. Everything is laid out in a logical fashion, and easily understood terms explain the various monitoring and management capabilities. eG VDI Monitor uses a browser-based management console that I found easy to navigate. The primary interface is broken up by tabs and submenus, while all monitored elements are presented visually. With a quick glance, I could judge the health of the infrastructure, which is presented as a colored bar graph with green indicating all is well, orange for caution and red for identified problems. The interface allowed me to monitor activity in real time, or choose an event-driven view that shows what has occurred over a time line. While the console is attractive, what’s important is what it can quickly tell you and how it can assist you in solving a problem.
The console allowed me to drill down into specific monitored events, and I was able to see the relationship between all of the affected components. That’s handy when trying to track down a problem with an application’s (virtual) performance. The management screen is organized in such a way that it was easy for me to view all of the events associated with an application and how each element was performing or affected at the same time, a capability that eliminates finger pointing and speeds problem resolution, sometimes before the end user is even aware of the problem-proactiveness at its best.
The console also enabled me to determine measurements at a glance, with its list of monitored elements, ranging from CPU utilization to disk usage to IP traffic, colored to indicate status: red means you have a triggered event warning, while green means all systems are go. This is the whole ideology behind the product-with a glance, you can determine whether everything is OK or what troubles may arise in the near future.
The reporting module offers extensive “static” reports that can give insight on performance over a period of time, the number and types of alerts that occurred, and an activity summary. For most administrators, the reports will be a good way to back up how a problem was solved, show where the blame lies for a particular problem, or prove ROI of the product and VDI solutions.
The product features “triggers” throughout, allowing notifications to be automated and so eliminating the need for someone to have to constantly monitor the console. I was able to create triggers (for example, I created an alert for overutilization) and then have alerts pushed out to via an e-mail (text messages are also supported).