Your VMworld 2011 Enterprise Virtualization Performance and Capacity Management Short List

With VMworld 2011 around the corner whose booth’s should you visit? Well if you are an enterprise with a large scale virtualization initiative then performance and capacity management should be at or near the top of your mind.  VMware clearly agrees as they have announced and delivered vCenter Operations specifically to meet this need. So after you go see vC Ops Enterprise, what else should you go see? 

VMware clearly intends to position vCenter Operations as the next (but probably not the last) management console that enterprise vSphere customers should buy. The principle need for a second management console is that the management of the performance, capacity and configuration (and their interrelationships) of the virtual infrastructure  is a sufficiently complex problem that is sufficiently different from the administration of the environment. This problem is also directly related to the performance of the workloads (the applications) which means that the user’s of a product like this are not the administrators of the virtualization environment, but rather typically the tier 2 and tier 3 support engineers who live in between the virtual infrastructure and the applications owners.

So let’s say you buy into all of this and conclude (as you should) that managing how the resources in and the configuration of your virtual environment impacts workloads justifies the purchase and subsequent maintenance and operation of a solution designed to meet these needs.

Enterprise Virtualization Performance and Capacity Management Criteria

When you start to evaluate these solutions you should carefully consider the following criteria:

  • Which virtualization platforms does the product support (and how will this change over time)? The new vRAM pricing has caused many enterprises to look at other virtualization platforms for their less demanding workloads. If you are going down the N hypervisor road, then you should absolutely pursue a cross-hypervior management stack strategy.
  • Is this an enterprise scale solution? Will it work for large environments, spread across multiple geographically distributed data centers? Will it scale to 10’s or even 100’s of thousands of VM’s? Can the product be effectively managed at this scale? Does the vendor know how to work with and support a global enterprise?
  • Can the solution monitor the physical environment that supports the virtualization platform (directly, and not just through the information provided by the hypervisor)?
  • Is storage performance, and in particular the impact of storage latency upon workload performance monitored and surfaced on a per host and per VM basis?
  • Are Capacity Planning and Management features included that allow you to avoid day-to-day resource bottlenecks and prevent running out of capacity in the future?
  • Is Configuration Change monitoring included, and is this cross-correlated with resource based bottlenecks and performance problems?
  • Is the notion of end-to-end infrastructure latency present in the product?  This is a critical way of understanding the performance of a virtual infrastructure as it looks at how long it is taking for things to get done instead of how much resources is being consumed.
  • Does the product include automatic, self-learning analytics and baselining? If this is included, how hard is it to set up and how hard is it to expand the scope of the data being fed into the system over time?
Quick Review of Vendors on the Short List
  • VMware vCenter Operations Enterprise. vC Ops Enterprise is a bundle of vC Ops Standard, vC CapacityIQ, and vC Configuration Manager, all glued together by the analytics that came with the Integrien acquisition announced a year ago. The analytics bubble the thousands of detailed issues that might or might not be real problems up into Workload, Health and Capacity scores. Unfortunately at this point CapacityIQ and Configuration Manager are not fully integrated into vC OPS yet, which means that vC OPS Enterprise today comes with three consoles and three databases. Enterprises considering vC OPS Enterprise should also drill deeply into what third party adapters exist for gluing existing systems management tools into the self-learning analytics and what professional services are required to make this happen.
  • Netuitive. Netuitive is the third party alternative to vCenter Operations Enterprise from the perspective of self-learning analytics with adapters into pretty much anything and everything that you might already own. Enterprises comparing vC OPS Enterprise and Netuitive should focus upon the ease with which third party data sources can be integrated, and how automatically the self-learning analytics adapt to both new data streams and changes in normal data stream behavior.
  • NetApp Insight Balance. Formerly Akorri BalancePoint, was the first end-to-end infrastructure performance management solution and is still the only solution that can map every VM that talks to storage to the spindle on the array that is serving that VM and then calculate end to end infrastructure response time from the VM to the array and back again. Insight Balance also includes some very sophisticated latency and load based capacity analytics that provide a different view of capacity and performance than other solutions.
  • Reflex Systems. Reflex Systems is a new vendor in the performance and capacity management realm with recently added capacity management functionality. The reason for considering Reflex is that it is built upon an extremely mature and highly scalable virtualization management platform that has the most robust configuration management capabilities in the industry.
  • Quest Software vFloglight. There are more VMware customers running vFoglight in production than there are customers of any other virtualization performance and capacity management solution. vFoglight is not only a market leading solution, it benefits from being a member of the full Foglight family at Quest. This means that it is easily extended with monitors (Quests calls these cartridges) that  go down into Storage, down into the physical network and up into the applications layer.
  • Veeam nworks. Veeam nworks is unique in that it comes in versions that that allow for you to use either Microsoft SCOM or HP Operations Manager to robustly manage your VMware environment. So if you already use one of these two enterprise management consoles, then Veeam provides you with a plug-in that extends these products into detailed managers of vSphere Performance and Capacity.
  • Virtual Instruments Virtual Wisdom. Virtual Wisdom is the only solution that can give you a real time, comprehensive, and deterministic view of every transaction that is flowing through your fiber channel SAN and how those transaction completion times are impacting the performance of physical and virtual workloads.
  • vKernel vOperations Suite. vKernel made its name in Capacity Performance and Management, but has expanded the functionality of the product with performance monitoring, optimization, and chargeback features.
  • VMTurbo. VMTurbo in unique in that they have a free product that includes many of the basic performance and capacity management features that other vendors charge for and that their paid for product is focused upon prioritizing workloads, and then automatically adjusting the resource allocation in the underlying environment to assure the performance of workloads in priority order.
  • Xangati VI Dashboard and VDI Dashboard. Xangati got its start by deeply analyzing Netflow data from within the physical and virtual networks that support a virtualized environment. This view of how the network is impacting performance is being complemented with deeper insights in storage being announced at VMworld. Therefore at VMWorld, Xangati will be the second vendor of a complete Infrastructure Performance Management solution.
  • Zenoss Service Dynamics. If you were going to build an enterprise management framework from scratch to address your existing physical environment, virtualization, private clouds, public clouds, cost of adoption, cost of ownership and easy expansion of functionality you would build Zenoss Service Dynamics. If you are thinking about tossing out your legacy framework (or at least banning it from your private cloud) start by taking a look at Zenoss.


This post left a lot of solutions off of the short list. The enterprise management frameworks were left off because they are too complicated to install, to complicated to maintain, to difficult to use, and generally too expensive to be attractive to the enterprise virtualization team focused upon finding an appropriate solution in this space. Many solutions from smaller companies were left off as well. In most cases this was because these solutions might be fine for an environment of 500 VM’s or less, but they do not scale up to larger enterprise environments. The vendors on this list are also the ones that come up most frequently in conversations with enterprise customers looking for solutions in this space – indicating a degree of commercial success for these vendors and their products.