As I was reading Mark Vaughn’s latest SearchServerVirtualization.com article, Virtualization management tools: Navigating the muddy waters, and Kendrick Coleman’s Next Generation Management - The Battle Begins, covering the release last week of vOPS Performance Analyzer, I was struck by how confusing the virtualization management tools research process is becoming. There are an increasing number of choices, major differences in philosophy and approach, and a massive amount of information that must be digested to truly understand these differences.
Mark Vaughn’s article highlights the high-level dichotomy between choosing virtualization management solutions from hypervisor vendors vs third party virtualization management providers. The main differences he notes aside from vendor sprawl are that best-of-breed tools are more likely to robustly support multiple hypervisors, but because of the speed and complexity involved in hypervisor advancements, third parties may not have enough visibility into the process to develop thorough support for all of a hypervisor’s features. This can lead data centers to “[manage] at the lowest common denominator”.
While this article does bring up a good starting framework to evaluate solution vendors, the recent history of management tool releases from VMware compared to those of third party vendors shows that these conclusions do not always hold. I’ll use VMware as a case in point for a “hypervisor vendor” as they have most recently made forays into virtualization management solutions.
Let’s look at two areas. First, in today’s post, we will review the VMware roadmap for products they have delivered in virtualization management. Tomorrow, we will examine integration challenges that a hypervisor vendor has when incorporating technology from acquisitions into their product line.
VMware is a large public company producing at last count from their website, 41 products spread over 9 solution areas. The products that put them on the map and fuel demand for their other applications are their hypervisors. Without advances to this critical product area, the others fall flat. As a result, it stands to reason that the majority of development efforts would be placed on hypervisor work, and that other products would receive less attention. This hypothesis is supported by the last 24 months of tech history when comparing the release schedules of VMware ESX against CapacityIQ (VMware’s capacity planning tool) as shown below.
This chart shows that for the past two years, VMware ESX has produced noteworthy releases at a regular bi-yearly pace for the most part. VMware CapacityIQ has on the other hand gone over a year between releases, most notably, forgoing several months of management support for ever-advancing hypervisor features.
Third-party vendors, whose main focus is on management use cases, will innovate swiftly to keep up with changes to the hypervisor and develop ever more compelling features that enhance what a virtualized IT organization can do. This rapid innovation is a virtualization management solution’s source of competitive advantage and is reflected in aggressive release schedules that make features available to end users as quickly as possible (VKernel’s release schedule in the past 24 months is used as an example in the figure above).
Another interesting point here is in how effectively a hypervisor vendor’s management tool developers are able to leverage internal communication with the hypervisor team to better support the hypervisor’s features. Recent technical history has again shown that for products not receiving sufficient attention, the potential advantage from unfettered internal knowledge sharing is negated. An example of this is that CapacityIQ in its first release did not support vSphere, the most up-to-date hypervisor release at the time, and continues to experience compatibility issues with in-house products to this day (VKernel on the other hand supported vSphere immediately).
So, just by looking at roadmaps and delivered products, third party virtualization management providers such as VKernel, appear to have an advantage in quickly addressing hypervisor advancements and user needs. But, what happens when a company buys their way into a space instead of developing a solution in-house?