In two weeks the VMworld 2014 Europe conference will be held in Barcelona. The whole ITQ vUnit will be there and even a couple of non virtualization guys will be joining us. At the VMworld USA there were a couple of announcements which I hope to learn more about at VMworld Europe. I also spotted a couple of new technologies and trends which could be easily missed if you focussed on all the Air, vRealize and EVO news. Here are a few of them which I’m eager to learn more about at VMworld europe 2014


You can’t have missed this one honestly. Docker is all the rage now and I personally think it will grow mainstream pretty fast. So docker itself is not the big news here. What makes it interesting is that VMware announced that they will be collaborating with Docker. This will probably be used in their Air offerings, but it would make sense to also dockerize applications which are shipped to customers.
Next to using Docker for VMware products I expect that VMware will encourage the deployment of dockerized customer applications on top of VMware virtual machine. VMware is already working on a new technology which makes using application containers the perfect way to scale an application:

Project Fargo

Previously called VMfork. And that exactly describes what it is. It is a technology that forks a running VM. You can spin up a running clone of a VM in just 500 milliseconds. So that’s not a disk clone, it’s a clone of a running VM. So no boot time, no copy time, just 500 milliseconds to create a running VM.
Of course that VM now needs an application to run. Having containerized applications suddenly makes perfect sense. Especially if you integrate the forking with a container clustering product like Google’s Kubernetes. Need to scale out? Spin up a new instance of an OS, tell the cluster manager it’s there and the cluster manager will start a container. Done.


VMware APIs for IO filtering. This is basically an official API for what PernixData is already doing with their FVP product. It allows 3th party vendors to hook into the storage path of a VM. Then their product can accelerate the I/O’s to flash for example, or replicate it to another host. That are exactly the use cases which will be supported when the API is released. VMware has been working with SanDisk on an acceleration product and with EMC on a replication product (Recoverpoint). The interesting thing is that somehow there are already vendors delivering products doing exactly what VAIO enables every vendor to do. I already mentioned PernixData but in the replication marked there is Zerto who is doing this already for a couple years. Makes me wonder why the big companies (SanDisk/EMC) who have all this R&D budget couldn’t figure out a way to do what PernixData and Zerto are already doing without any official API.


To summarize all of this: It’s all about decoupling. Docker decouples the application from the OS, Fargo decouples running VMs from deployment times and VAIO decouples storage features from the storage hardware. Or to put it in marketing terms: Everything is moving towards the Software Defined Datacenter. Or was it called the Cloud?

Christiaan Roeleveld

Author Christiaan Roeleveld

I was introduced to virtualization in 2004 and have grown addicted to the technology ever since. I like working with new technology, expand the possibilities and use it to the fullest. The last couple of years I have seen the power of infrastructure shift into software, the hardware has become less important, software is key. Software Defined will make the automation potential at the infrastructure level nearly endless. Below you can read my blogs, a personal automation blogsite can be found here .

More posts by Christiaan Roeleveld
1 October 2014

Leave a Reply