Lately I have spent quite a bit of time talking about and defining different cloud terms. But the last few weeks have provided me with some additional perspective in terms of what people are looking for from “the Cloud.” They don’t just want “the Cloud” – they want the ability to manage the cloud and be in control of the data they put there and that inability to do so is what still gives users pause about “the Cloud.”
In talking to a number of users over the past few weeks about “the Cloud” and what they want from it, it is clear this is the strategic direction they want to take their organizations. But the feedback they also provide is that “the Cloud” is more hype than reality in their environments and that they have to progress very slowly and cautiously down this path for a variety of reasons that include:
- They want to control and manage “the Cloud.” While some companies are ready to turn over the control and management of some of their data to a third party, most are not. IT administrators and directors want to retain and feel like they are always in control of their data but as they move it to the cloud they still feel like they are giving up too much control. The last thing they want to feel is powerless should something go wrong and right now cloud providers are not yet giving them that assurance.
- They want a single portal from which to control and manage “the Cloud.” It seems that this portal is becoming vCenter. IT staffs are not growing (though I am getting conflicting information on that point) so organizations need to continue manage more with the same people. To facilitate this ideal the guys who support the infrastructure do not want to use multiple portals (vendor provided or other) to manage their cloud infrastructure, they want one. Since most organizations started virtualizing their server infrastructure with VMware and are using vCenter to manage their VMs, they want their cloud storage infrastructure to plug into vCenter so they can manage their storage from that portal as well.
- They first want to build clouds in-house. Before they hand off any of their business critical applications or application data to a third party, they are building private clouds in house. To a certain degree, this defeats one of the purposes of a Cloud model. They still have to lay out CAPEX dollars to build this cloud-enabled infrastructure instead of going directly to an OPEX model that theoretically provides it. I attribute this decision to two factors. First, most storage vendors are not setup at this point to support an OPEX model and, second, users view it as less risky at this point to build their own cloud as opposed to hosting it with someone else.
- The storage buying decisions are becoming very complicated. It used to be when buying storage arrays the decision pretty much boiled down to what volume management features you wanted on an array (LUNs, LUN groups, RAID options,) how many FC ports you wanted and how well it performed. Now storage array buying has become a highly complex exercise. Those other features are still part of the equation but features such as sub-volume tiering, thin provisioning, snapshots and replication as well as integration with data protection software are all part of the discussion and must be reviewed individually as all of their features may be used in a virtualized environment to deliver on the ideal of the cloud such that it operates as a utility and applications and their data are constantly available.
Users like the idea of “the Cloud” and are gravitating in its direction but they are not willing to take a leap of faith in the hopes that a cloud provider will be there to catch them should they fall. Instead they are taking a much more measured approach toward implementing the cloud giving preference to those solutions that enable them to implement the cloud in such a way that they can manage and control it.