The vRAM was the license model VMware used in vSphere 5.0. It basically limits the usage of virtual memory, which is different from physical memory, per license. When first announced last year, it created a lot of angry customers overnight even though VMware estimated that the license scheme wouldn’t affect most of the existing customers. Later on, VMware doubled the amount of virtual memory and implemented a cap per license, and insisted to roll out the modified license model despite strong objections from customers.
One year later, at this year’s VMworld 2012 keynote then “almost” CEO Pat Gelsinger announced that VMware would abandon the vRAM license. In fact, about one week before his official announcement this CRN article leaked the news already. Like most IT rumors, the article turned out to be quite true.
Bothered by SLOW Web UI to manage vSphere? Want to manage ALL your VMware vCenters, AWS, Azure, Openstack, container behind a SINGLE pane of glass? Want to search, analyze, report, visualize VMs, hosts, networks, datastores, events as easily as Google the Web? Find out more about vSearch 3.0: the search engine for all your private and public clouds.
Interestingly, Pat mentioned that other VMware colleagues told him that there would be standing ovation upon the announcement. I was sitting at back and saw only two standing up in front of me the second time. Think about it – if a vendor made a mistake and then fixed it, things just get back to what the way it should have been in customers’ eyes. I can’t imagine that the vendor expects standing ovation, instead of asking for forgiveness. That seems to me that VMware team had not yet thoroughly reflected the whole thing. Without such a thorough reflection, similar mistakes might happen again in the future. That is what concerned me and possibly other customers.
What Are Damages?
Although the vRAM is recalled, the damages are there already. It’s pretty hard to estimate the size of the damages, but I personally think the damage could be one billion dollar or more.
For one thing, most customers including me got confused about the vRAM especially how it’s calculated and accounted across a vCenter in the first place. Therefore, VMware marketing spent lots of efforts to explain how vRAM works, and more importantly why most customers wouldn’t pay more with the new model. As usual, people mostly choose not to act or to act slowly when confused. By that VMware might have lost some opportunities due to the new complexity.
Even worse is the market share. Competitors like Microsoft tagged vRAM as vTAX. All of sudden, a new topic became important: how to save with vSphere if you still love VMware? With that, Microsoft made significant advance in marketing its Hyper-V.
The biggest damage is, however, that VMware lost the unique trust from customers who used to think of VMware as a different vendor that they can fully trust and follow. With the forceful introduction of vRAM, VMware became yet another vendor, not much different from others like Microsoft, Oracle.
I know many customers who had standardized on VMware started to look out for alternatives. This should be a conventional wisdom, but only upon this event did most customers realize that hypervisor diversification should be treated as a priority.
What Was Behind The vRAM Decision?
Now with all these obvious problems, why VMware insisted on vRAM? For sure, VMware does not hire dummy employees. But why smart people made the obvious mistake?
All that rooted from the deepest fear of loss of revenue with advancements of modern CPUs especially those with more cores. If the CPUs continue to double its processing power every 18 months per Moore Law, customers would need much less licenses down the road if they just need same amounts of workloads. That of course directly translates to reduced license revenue for VMware.
What is the best way to measure workloads? VMware has plenty of technical talents who knew exactly the best answer. That is why vRAM got selected. I think it’s the best way to address the revenue concern from a technical perspective.
But the license model is essentially a business issue, not a technical issue. Technically, you can support any license model you can imagine. The decision should be made by marketing team. In this case, it should be a top executives’ call in the end given the importance of vSphere which is the only cash cow at VMware.
The vRAM is over but the lessons stay. I think here are a few lessons VMware and other vendors can learn from this:
- Customer satisfaction is still important in cloud age. A vendor can keep growing business only when customers are happy. The only exception is in a monopoly market where there is no alternative. In virtualization market, VMware is still a clear leader but not in monopoly position. As one twitter message pointed out, it’s competition not the customers that reversed the vRAM. Very true.
- Don’t break a convention in a negative way. The industry convention is to license on CPUs. The same concern with multi-core CPUs is not unique to VMware at all – it applies with many other vendors like Oracle with its database. But no one wants to be the first to change it. Licensing is essentially an interest game between a vendor and its customers: one’s loss is the other’s win. Unfortunately, a vendor has to make up the gradual loss with volume, bundling of other software sales, and/or services, due to competitions.
- Keep business simple and consistent. No one likes complexity and much so when it comes to business. Even VMware abandoned vRAM, its competitors like Microsoft can still accuse VMware of inconsistency. After all, VMware did not make any promise that it won’t change its current model again.
Although the vRAM is abandoned, VMware will focus on bundling software in vCloud Suite to grow its sales. Will it affect VMware’s revenue in short term? If what VMware told customers is true that the vRAM doesn’t affect existing customers, it’s not likely. So before the bundling strategy is fully rolled out, VMware still has time to polish its strategy and product suites.