An interesting recent trend I’ve been seeing is that of new vendors differentiating their cloud service through giving customers ultimate flexibility around what they use on their cloud – rather than shoehorning them into a couple of OS options, or a few pre-defined server sizes and specs, these vendors provide an ultimately flexible buffet of options for their customers. The argument from these vendors, who generally target existing enterprise customers, is that when considering moving existing workloads to the cloud, an organization will want a virtual environment that perfectly matches their physical one – rather than one which is, in some cases, starkly different. This is the line RackWare is taking as it launches its “RackWare Management Module” an intelligence and automation toolkit that allows users to build their cloud (public, private or hybrid) without making any application or operational changes.
One of the key points RackWare is trying to ram home is that of mobility – as users face increasing concerns about being locked in to one cloud vendor, RackWare automates cloud mobility by supporting any type of infrastructure – not just cloud but also plain virtualized and physical. They also provide that holy grail, a push button automated mechanism to move existing workloads between data centers and on to cloud infrastructure. Another aspect of this mobility is elasticity and RackWare is promising seamless cloud bursting potential using its tool. RackWare describes it’s product line over three distinct areas as shown in this diagram:
Today, and into the mid-term future, there is an obvious need for tools that automate the wildly heterogeneous variety of resources as organization have a wide infrastructure footprint across physical, virtualized and all flavors of cloud. I suspect however that over time the tendency will be to manage traditional infrastructure with traditional tools – and lump all the virtualized and cloud infra into one pool This speaks to a core point of contention in the community – that of whether existing legacy workloads will be left in situ or organization will try and forklift them onto new infra. I tend to be in the camp that believes that, most of the time, forklifting will be just too problematic and organizations will leave legacy stuff where it is. That being the case, a product that is mainly differentiated by the mobility it brings to enterprise workloads will lessen in importance.
This is especially so given the promise that initiatives such as OpenStack and CloudStack (and, to a lesser extent, Eucalyptus cozying up to AWS) bring – the ability to move workloads between different cloud vendors, and between public and private, quickly and easily. RackWare concedes this point, in my briefing they admitted that their mobility offering will be commoditized over time, however they were adamant that tying policy and automation into the equation gives them an increased degree of stickiness. In their view they’re a “RightScale for enterprise” – that’s a valid differentiator given the landscape right now.
RackWare has a difficult job to do – on the one hand they need to sell enterprise on the idea of shifting workloads while on the other continue to articulate a message that reduces enterprise appetite to simply go “all in” on the cloud – that’s a difficult balancing act and one which introduces conflicts at every turn, it’ll be interesting to see how RackWare deal with that.