You might recognise the title in different terms. It’s very similar to “The King is dead, long live the King!”. As Wikipedia explains the original phrase was translated from the French Le roi est mort, vive le roi !, which was first declared upon the accession to the French throne of Charles VII after the death of his father Charles VI in 1422. In France, the declaration was traditionally made by the duc d’Uzès, a senior peer of France, as soon as the coffin containing the remains of the previous king descended into the vault of Saint Denis Basilica. So what’s this got to do with VMware vCloud Director I hear you ask?
Has VMware killed vCloud Director? What should you do if you have vCloud Director or if you’re thinking about implementing it? What does this mean for the vCloud Director based certifications such as VCP-Cloud, VCAP-CIA, VCAP-CID and VCDX-Cloud. I’ll give you my opinion of the answers to these questions in this article. It’s all about the Software Defined Datacenter.
This year, in August at VMworld USA in San Francisco, vCloud Director was pretty much completely missing from the event. Except of course my session on running Databases at Maximum Performance in a Software Defined Datacenter – VAPP4683 (Presented with Mark Achtemichuk in San Francisco and Andrew Mitchell in Barcelona). Many customers and partners were lamenting the death of vCloud Director. This wasn’t quite what VMware was doing, as I’ll explain during this article. Needless to say some customers and partners I spoke to at the event were not at all happy with the decision. Some customers had invested multiple millions of dollars in vCloud Director projects for Private Clouds.
For anyone that is not familiar with VMware vCloud Director, it’s a software solution that allows any company or service provider to implement a software defined datacenter, or at least a good number of components of it. It allows for self service automated provisioning of virtual machines and virtual applications into a private or public cloud. So you can deploy your applications whenever you want and within minutes they’ll be available. The best use case for private clouds in my opinion is to streamline the software development lifecycle and greatly improve the economics of software development projects. But it also has benefits for production workloads as well, as it takes away a lot of the manual tasks and decisions that administrators would have traditionally had to make when provisioning new applications. I have had customers that have reduced their project development and testing time by 50% using vCloud Director. In addition to reducing the test and development time customers have reduced their defect rate by orders of magnitude, which means their software is more reliable and they find many less defects after the solution has gone into production, which in term improves availability, performance and lowers ongoing maintenance costs.
During VMworld VMware announced a new strategic direction, and vCloud Director wasn’t it, at lest not for Enterprises. The new direction was vCloud Automation Center (vCAC) for Enterprises (private cloud), which was introduced with one of the best integrated demo’s I think I’ve ever seen during the main keynote on day two (called the Mother of all Demo’s). What this meant is that vCloud Director was for only going to be for cloud service providers (public cloud). As part of this change of direction VMware announced that vCloud Director 5.5 (which was released shortly after VMworld) would be supported through to Q3 2017 (4 years instead of the usual 2), and vCloud Networking and Security (Formally vShield) would be supported through to Q3 2016. Neither would be available for sale separately from September 2013, you’d have to have the vCloud Suite. A version of vCloud Automation Center is available and was added to every edition of the vCloud Suite, dependant on the edition purchased. So everyone with vCloud Suite, get vCAC.
vCloud Director is Dead?
So is VMware killing or has killed vCloud Director? No! Far from it. VMware is investing more in vCloud Director, but is focusing the investment on public cloud service provider requirements and the requirements of the VMware vCloud Hybrid Service (vCHS). This is why the title of this article if vCloud Director is Dead, Long Live vCloud Director. It will be around for a long time to come and will deliver ever more value to the world, with a slightly changed focus. VMware will be making the functionality from vCloud Director that is for Enterprise available through a combination of vCloud Automation Center and vCenter, some features of vCloud Director not applicable to Enterprise use cases will not be making the transition (primarily multi-tenancy and multiple authentication domains).
So does this mean you all have to rush out and implement vCloud Automation Center (vCAC) or migrate to vCloud Automation Center or that you have just seen your investment in vCloud Director flushed down the toilet? No! The change in strategic direction isn’t like turning on a light switch and everything is working. Just because marketing might say vCAC is the new way to go doens’t mean it’s actually reality today. Your investment is protected and you get support through to Q3 2017 (vCloud Director 5.5), plenty of time to get a massive ROI, and you’ll get bug fixes and patches during this time. This change in direction from VMware will take place over a number of years, and you will be given time to adapt, and the migration tools to do it. Why is vCloud Networking and Security support ending a year earlier that vCloud Director, I have no idea, especially seeing it’s an important part of a vCloud Director environment.
Right now vCloud Automation Center isn’t even feature equivalent to vCloud Director yet (based on version 5.2). If you’re using vCloud Director for self service development and test environment provisioning and automation, or self service catalog, it’s either going to be difficult, extremely hard, or impossible to get the same functionality out of vCloud Automation Center, at least the current shipping version and probably the next one also, especially if you don’t want to spend a lot on services (Remember Lab Manager vs vCloud Director 1.0?). VMware also needs time to create migration tools for customers to make the transition to the new way of doing Software Defined Datacenter for the enterprise, as there is no in-place upgrade path possible for vCloud Director to vCloud Automation Center. Getting the full migration strategy in place for customers is going to take time and it isn’t there yet.
vCloud Director isn’t dead, but what should I do now if I have it?
Firstly, don’t panic! If you’re an existing vCloud Director customer, whether the project has been delivered and is in use, or still in flight, I would highly recommend you continue down the vCloud Director path. Especially if your use case is self service catalog, self service provisioning of isolated test and dev environments, and you’re going to be using vCloud Director primarily for streamlining your SDLC. I would not recommend throwing in the towel with vCloud Director right now, it is an excellent solution for the right use cases and the right requirements. Having designed and implemented about a dozen private clouds and around half a dozen public clouds with vCloud Director I’ve seen the value it can deliver.
I would further recommend that when the next version of vCloud Automation Center (whatever comes after 5.2) is released that you consider implementing that along side but separately from vCloud Director for self service provisioning of production workloads, or in a test environment so you can at least get used to it. It goes without saying that if you’re not on vCloud Suite licenses you should get there as soon as possible, it’s the best solution for delivering a software defined datacenter and it’s incredibly cost effective compared to purchasing the equivalent licenses separately. vCloud Suite licenses will be a worthwhile investment, and will get you the optimal ROI.
What if I’m a net new customer looking at building a software defined datacenter with vCloud Director?
If your use case for vCloud Director is primarily for production workloads and provisioning and managing the lifecycle of production systems, and you’re a net new user (no existing vCloud Director), then I would recommend you work closely with VMware to plan an implementation of vCloud Automation Center as soon as the next version is available. The reason for this is there is no in-place upgrade path from vCloud Automation Center 5.2 to the next version, it will be a migration. So you might as well wait for the next version to become available if you can, or be prepared to stay on vCAC 5.2 for a while until the migration process is well defined and the tools are there to make it straight forward (or as much as it can be). I know a couple of customers that are on vCAC 5.2 and are happy with it for now for their use cases. Like vCloud Director, vCAC is a great solution for the right use case and the right requirements, and of course it’s part of the vCloud Suite.
If your use case for vCloud Director is making your SDLC, Dev/Test greatly more efficient and streamlined, and you’ve got no current vCloud Director environment, then I’d recommend you go with vCloud Director and the vCloud Suite. It can be implemented fairly quickly (with competent and skilled people), and you’ll get good use out of it till 2017. This’ll give VMware time to get vCloud Automation Center and vCenter sorted out to be feature equivalent (or close enough) to vCloud Director, and for the migration tools to become available and mature. You should work closely with your VMware partner and VMware to plan an eventual transition to vCAC, when the time is right, and once you’ve got your ROI.
If you’re not an enterprise, but a cloud service provider instead, then the choice and long term solution is easy, it’s vCloud Director. vCloud Director is here to stay and will be developed for the service provider market.
What does this mean for the VMware vCloud Director based Certifications?
The vCloud Director based certifications, such as VCP-Cloud, VCAP-Cloud Infrastructure Design, VCAP-Cloud Infrastructure Administration, and VCDX-Cloud will all continue. They will continue to have value to customers that are running vCloud Director, and they’ll have a lot of value for public cloud service providers. VCDX-Cloud will probably be valuable only to public cloud service providers in the future, or to VMware Partners that are providing consulting and services around private and public clouds. It’s likely that VCDX-Cloud will adapt to non-vCloud Director technologies, especially as vCloud Automation Center and OpenStack become more a part of the VMware landscape. I’m in the process of working towards VCDX-Cloud, and this change of direction by VMware has not altered that plan. I would encourage anyone that is working for a public cloud provider or is planning to work for a public cloud provider (based on VMware vCloud Director technology) to go through the vCloud Director based certifications.
There are probably many reasons why VMware has chosen this new path for vCloud Director Enterprise customers. vCloud Director was hard to integrate with the other VMware solutions, such as Site Recovery Manager. Partners were very slow to support backing up and restoring vCloud Director, especially the tenant objects (this can be done now with partner provided solutions). If you can’t easily and efficiently protect the data inside the solution it’s hardly suitable for an enterprise production workload right? The adoption rate was probably less than expected, but what do you expect when vCloud Director originally wasn’t feature equivalent to the product it was meant to replace (Lab Manager), had no migration path, and had no training for a long time. vCloud Director was complicated and had lots of moving parts, was hard to understand (I don’t agree with this, but I know some people have this opinion). Maybe many people didn’t understand and couldn’t articulate the value proposition for vCloud Director as it was very different to a traditional infrastructure or virtualization solution. What business problems was it actually going to solve, what was the ROI and what were the benefits (hint: refer to the start of this article for some)?
Whatever the reasons are I think this change is a good thing. It will allow VMware to deliver solutions (over time) that solves the Enterprise and Service Provider use cases, is simpler and more automated, more focused. For Enterprises the solution will have much easier integration into the wider VMware product stack. For Service Providers they’ll be able to more seamlessly integrate with their customers, be more scalable, and have significantly more easier and less disruptive cloud maintenance. It will be easier to deliver the Software Defined Datacenter Vision and Hybrid Cloud. All the while protecting existing investments and providing massive value.
This post first appeared on the Long White Virtual Clouds blog at longwhiteclouds.com, by Michael Webster +. Copyright © 2013 – IT Solutions 2000 Ltd and Michael Webster +. All rights reserved. Not to be reproduced for commercial purposes without written permission.