This week has brought two fascinating and contradictory perspectives about how large IT organisations view cloud computing, and how they're responding to the messy muddle that is today's cloud computing environment.
The first article is from the UK version of CIO, which discusses the recent Intel-sponsored data centre consortium, and then quotes a number of IT executives regarding cloud computing. The second article is an interview with Michael Heim and Mike Meadows, CIO and VP, respectively, of Lilly, a large pharmaceuticals company.
The UK article is headlined: "Is the cloud just marketing hype? CIOs have their say." Using the Intel announcement as a jumping-off point, the article notes that many CIOs are overwhelmed by vendor noise in the cloud computing space, and suspect that many vendor's claims are merely "cloudwashing," or, put another way, old wine in new bottles labelled "cloud computing."
One IT executive is quoted as calling for cloud standards: "a body of common, industry-wide standards are going to be crucial." This executive appears to be taking a wait-and-see attitude, reluctant to move forward until cloud computing is available in common, standardised offerings making it easy to mix-and-match.
It's understandable that he would react in this fashion. One commonly hears a lot about the fear of lock-in, of choosing a cloud provider and then being chained to it, even if vendor or user circumstances change, making the choice no longer desirable. The goal of this approach is to ensure that the user has many fungible choices and can leverage price competition among suppliers. Given the high profile of this type of reaction, it would be easy to conclude that user organisations are waiting for the dust to settle among cloud providers, allowing an easier decision process once stability in the market has occurred.
Lilly's take on standards
Turning to the Lilly interview, a completely different perspective is evinced. Lilly has had a fairly high profile regarding its use of cloud computing. It's pretty well known that it has used Amazon Web Services for large scale number crunching. Heim's interview makes it clear that Lilly is continuing to move forward and is planning to increase its use of cloud computing. The highlights of the interview, with some comments:
Cloud computing starts with a business unit: Heim notes that it was a research group, not IT, that began experimenting with AWS. This aligns with the common experience that many early users of cloud computing are business unit users driven by pressing business goals or opportunities. This is not to say that this use is trivial, as many cloud computing detractors describe early adoption patterns. Pharmaceutical number crunching is a big deal with heavy compute use and lots of dollars at stake.
IT gets involved to standardise use patterns: Heim says that IT got involved to create standard templates, thereby making it easier for cloud computing users to become productive more quickly. That these templates standardise use patterns and "control" end users is a side effect of the fact that they accelerate use and increase productivity.
Systems management is the crucial linchpin of Lilly's future: Lilly plans to support both internal and external cloud environments, with multiple types in both arenas. Meadows identifies systems management (i.e., managing cloud environments) as critical to support its plans, because it wants to have one overall management infrastructure which hides the operational differences among the various offerings. There's a lot of loose talk about hybrid cloud environments around today without much awareness of the challenges of such an environment, and Meadows clearly identifies the key element to implementing a hybrid approach: a common management framework. It's worth quoting him to understand this reality:
"Our strategy is we don't believe we will place a heavy bet on any one vendor of cloud services. It will be utilised in the full spectrum of cloud services, whether it's fully external, private to us or a hybrid of the two. That puts a tremendous criticality on the middle layer of systems management to operate those, so that we don't need to know the specifics of every individual cloud provider. We're spending a lot of focus with a central team on how can we, using providers who are responsible for that or have that as their business model, or our own work, generate a middle layer to allow us to use multiple models and interoperate between the two, private and public."
Cloud computing offers real financial benefits: And it's not just for transient use cases, either. Meadows notes that Lilly is running financial scenarios and finding that public clouds offer savings even when used for full time computing.
The discussion then moves on to what, to my mind, is the most interesting part of the interview: the question of lock-in. Heim recognises that lock-in is a concern, but says that Lilly also focuses on what he terms "time to value." Here is his quote:
"I think the attractiveness in the model is the time to value. We're always concerned about lock in and exit strategies, but we need to stand on the other side of that and talk about the time to value, and what's the value that's generated by getting in there early and leveraging the model."
In other words, the flip side of waiting for standardisation is opportunity cost, the lost financial benefits incurred by waiting.
This opportunity cost is what many people fail to grasp about deferring decisions: there's a real consequence to not doing something, just as there's a consequence to taking action. In terms of this debate, waiting in order to avoid lock-in imposes a different cost, but imposes a cost nevertheless.
Of course, it's hard to recognise, let alone calculate, opportunity cost. By definition, it's unknown. It's far easier and safer to identify the cost of lock-in. But one shouldn't characterise lock-in as something imposed by vendors upon hapless users. No doubt many vendors exploit customers who are committed to the vendor's technology.
It's important, however, to understand that many organisations embrace lock-in as the inevitable companion to achieving time to value. Lilly clearly believes that the rewards it is achieving by courting lock-in are worth the downside that accompanies that commitment.
Looking over today's cloud computing landscape, it's obvious that (to mix a metaphor) it's fog-shrouded, making it difficult to chart a safe course. Much of that fog is generated by vendors who are trying to disguise their tired old products to make them look more relevant to cloud computing (no doubt by substituting the phrase "cloud computing" for "virtualisation" in their brochures, performing a "re-washing" of the last disguise attempt).
In such an environment, it's tempting to capitulate to confusion and defer taking action. Casting that delay as "avoiding lock-in" and "ensuring standard conformance" sounds reasonable and judicious, but it's not. Instead, it's a formula for being bypassed and relegating oneself to the second (or third or fourth) division.
The "waiting for standards" watchword is particularly pernicious. Standards make sense when all the innovation in a particular technology has been wrung out, and standardising it allows innovation to take place in some other area that leverages the low cost and wide availability of the now-standardised component. We are far, far from that state in cloud computing. To paraphrase Winston Churchill, we are not even at the end of the beginning of cloud innovation, much less the end of it. There is going to be enormous development in the characteristics, capabilities and varieties of cloud computing over the next five years, and pausing until it's standardised is futile.
In an environment of tremendous change, with significant financial benefits available to embracing that change (a la pursuing "time to value"), a better decision is to make a choice, even while recognising that it's impossible to know if it's the "right" one. One thing is for sure, when chaos is swirling around you, sitting tight is not is not a safe choice.