We recently had the pleasure of having an extended conversation with Antonio Piraino, CTO at ScienceLogic. Our chat covered a wide range of topics, including the cloud industry at large, enterprise adoption of cloud, and the differences between the developments in cloud technology and the developments among cloud service providers.
Gathering Clouds: What are the major industry players doing right and what are they doing wrong, both on the MSP side and on the technology development side?
Antonio Piraino: So I’ll start with the service providers: I think that AWS is doing a great job in many ways. It’s evident from the revenue that they’re generating in the US, but I think they’re going to start hitting hurdles, because they don’t have a support mechanism in place to really be, “high touch.” From a US-centric perspective, I think they’re doing it right in terms of innovation. But in terms of global market penetration, they are going to have a tough time.
The service providers are doing it right because they’re introducing so much automation into their technology. And you can see the results of that as the revenue from each employee gets higher, as do the margins. If that’s your threshold of success, which I think it ought to be, then you have to say, “Well, who’s doing it wrong? Who hasn’t deployed it in that way?” And that would have to be RackSpace and I know that sounds like sacrilege, and I like them a lot. I like the individuals in the company a lot. But going forward, they will be viewed as doing it wrong if they don’t fundamentally change a lot of things inside their organization, in terms of the volume of people that they seem to continue to throw at every problem or request or testing situation.
On a technology side, I think the VMware’s and the CISCO’s of the world are obviously doing it right. And they’re doing it right because they’ve been rewarded through increased expenditure and acquisition on those technologies. With the tools that they’re both bringing out and the converging infrastructures that they’re bringing out that are really working, and everybody else is trying to follow them now.
The organizations who remain inefficient, who remain very people-oriented and companies who are not fixated on creating more efficiency and better tools in this battleground are going to lag behind. Even when we moved from mainframe to server client, there were a bunch of tools that dropped away. Now, as we move from server client to cloud computing, there are even more tools that we are starting to see drop away. Many companies competing in this market are just not able to handle the level of multi-tenancy and scale that cloud computing brings. All the big companies that you’ve heard about ten years ago are not necessarily going to be the winners of tomorrow.
GC: How do you express the value of cloud today? And how do you see that changing in the near future?
AP: The value of cloud today is really in the promise that it offers. It offers you an out if you realize that you have great inefficiency in your business operations because you now realize that you can start to rely more on IT while also having more options available at your fingertips. Not everything is priced the same and you don’t have to utilize the same technologies or the same service provider for all of your workloads. That’s where we start to enable this broad term of “hybrid,” topologies. I greatly respect the enterprises that have figured out the benefits of a hybrid environment because they realized what workloads belong in a public cloud, what workloads belong on-premise; what workloads belong in a highly managed environment and so on.
When we talk about cloud computing, there are a lot of enterprises out there who mistake virtualization for cloud computing; virtualization alone does not a cloud make, after all. With virtualization, sooner or later, you reach a point of diminishing returns if you’re relying purely on virtualization to bring this “wow” factor to your boss or your CIO because you’ve reduced your IT footprints, or consolidated some costs. In 2 or 3 years’ time, you’re going to be right back where you were from a footprint perspective, and probably even greater on the cost side of the equation because you’re going to have more saturated servers, bigger servers, and more utilization of the virtual machines. And your costs will again start to go up. And I’ve done studies of these situations in the past that have shown that once you’re virtualized, your costs actually start to increase. Your debt, energy consumption – everything increases.
The real focus of the industry is going to be around two core areas. First, in the immediate term, the focus will be on control and manageability of this fluid IT environment which is constructed of highly distributed applications on numerous devices, including mobile devices. The second area of focus will be on a higher-than-ever utilization of data to make smart operational and business decisions. Cloud and big data are attached at the hip because ultimately the “internet of things” relies on the fact that you’re going to be making much smarter decisions in your business operations and your IT operations, so you’re going to require really smart well-correlated data points to make those decisions. Now where does cloud computing come into that? If you don’t have a robust, underlying cloud computing platform as a baseline to enable those capabilities, to enable big data usage, to enable the manipulation of the data, to enable the archiving, to enable the delivery and correlation, then you never really get there.
Now, I do think that the underlying cloud platform has become very commoditized, but they still remain the underpinnings of where computing will ultimately go. Cloud computing in five or ten years’ time will be discussed at a data layer and an application layer, at a mobile device layer, rather than at the infrastructure layer. But you’ve got to get that baseline correct. And I think that’s where cloud computing is going.
To that end, when one considers “What is missing in the cloud industry?” I think, even though as many people talk, myself included, about automation and convergence, there’s a long way to go in terms of optimizing the level of efficiency and correlation in decision-making between the Network Operations Center (NOC), the Security Operations Center (SOC), the systems people, the facilities people, the networking people and so on. ScienceLogic, for example, has the capabilities of helping to link many of these groups together and drive a lot more of that correlation and automation. But the companies, service providers and enterprises, have to get themselves to a level of confidence in the internal policies as well so that if they apply policies against some level of automation, they will feel confident that it’ll be carried out in an efficient way. That means creating a culture that understands the value of breaking down the walls between these groups. When all of that happens, with the culture changes and everybody embracing automation and efficiency, then organizations will really start leveraging cloud computing to do much smarter, more decisive things, with the data that is produced. That’s where we enter this world of “the internet of things” which we‘re still years away from. It is a much longer road than we think, in terms of the automation and convergence.
GC: From your perspective, what really characterizes the difference between the Europe and the US cloud markets?
AP: Well, I know this sounds like a very obvious answer, but probably the first big difference is that Europe is made up of multiple countries and borders, and has much better cheese than the US… so, that’s my joke for the day.
Though having multiple countries and borders leads right into the discussion around regulation. That is a fundamental difference to the US: the fact that they might be of similar size and population (though much smaller in land mass), but having all those borders is a completely different kettle of fish. It means, for example, AWS is now precluded from playing quite as aggressively across Europe because they don’t have the same number of data centers there as they do here. Unless they’re willing to completely ignore the rules and manifest similar to how they play in the US, they will remain smaller in presence in Europe.
Germany has very stringent rules on data, and yet some major companies completely ignore those laws, moving data across borders at will. This issue, for example, leads to another major difference in the European Union compared to the US market. US polices are much more stringently enforced and people play by the rules a lot more. Frankly, Europe doesn’t quite do it as well, on a holistic basis, instituting regulations and then enforcing them.
In Britain – and this is my own takeaway after years of watching how the Google’s of this world were sued left, right and center across Europe – a lot of the things that tech companies get sued for is related to privacy concerns. People don’t want photos of themselves and/or their house being put on the internet. Meanwhile, when it comes to Big Brother in Europe, they seem to be quite embracing of it and quite accepting of it. In the US, it’s exactly the opposite! I would say that the people are quite happy to have their faces all over Facebook and their house on Google Maps, they’re quite proud of it, yet, but God help us if Big Brother‘s watching me. It is a very different mindset.
I know that in the United Kingdom (UK) for example, some of the customers we have there are concerned with the Patriot Act in the US, which typically governs all US companies no matter where they are in the world. However, in the UK, they have their own set of regulations that already protect privacy of data for individuals as well. My point is that there are a lot of fights where regulatory vs. data issues are concerned and ways to go in terms of sorting out the issues between the various country borders.
Even from a security concerns perspective, from a data retention perspective, from a processing of data perspective, there are many different rules in Europe and then suddenly the EU comes along and tries to make strong recommendations that supersede all of these different countries’ rules and regulations. Frankly, I don’t think that any of them are that meaningful because each country is full of very sovereign states and they’re going to really battle over whether to listen to the EU Commission when they’ve got their own sovereign rules in place. There’s a way to go to get around all the bureaucracies and process governing cloud adoption in Europe. I know we hear cloud is growing in Europe but I actually think that it’s perhaps growing from a mobile perspective faster.
Now I don’t mean to get tangentially into the definitions of cloud, because I think it has a very specific definition but if we are being very liberal with its usage, I think a lot of the cloud is being dictated to from mobile device usage, and Europe has always been further ahead than the US in terms of mobile applications and mobile device usage. In the US, adoption seems to be driven much more at the infrastructure layer and up. So I think there are a number of differences in the regions. The size of the companies also makes the difference which means that they’re far more sensitive to prices in Europe than they are in the US, which I believe matters more than anything else in Europe.
If I look around Europe, Britain and Scandinavia seem to be more permissive than other European nations about the way in which physical data can be processed in the cloud. Places like Spain and France impose much tougher requirements such as requiring cloud service vendors to know at all times where their information is being kept. In Germany, as I previously mentioned, they have very strict laws on data in the cloud, but it’s very laxly enforced. Switzerland is a country unto itself as well, and now it’s becoming known as a big data residence, particularly from Russia and a lot of the Eastern Block countries.
Each country has a long way to go before they have a genuine European cloud computing paradigm like we do in the US. I think they have national concepts, which is quite helpful for anybody who’s innovative in-country, but less helpful to the AWS’ of this world.
GC: So what do you attribute to AWS’ dominance there not being as great as it is in the US?
AP: I would say that obviously the regulations and borders are more complex. But more critically, AWS just doesn’t have the following or the mind share that it does in the US, both from the consumer perspective and then also from the corporate perspective as well. In the US, we’re used to clicking on Amazon Prime now, and it’s a short hop from there to AWS. It’s not in the everyday lexicon in Europe, dramatically less in the UK and even less so in continental Europe. AWS is not an easy access point for the average small business. I still think it really takes a techie or a developer to use it. Strictly speaking, AWS is not geared for accessibility because it’s just not easy to use. It’s really a toolkit, rather than a nice friendly hosting environment. Also, with such a small data center footprint in Europe, you’re limited in terms of how much core data and data processing can take place on these cloud environments and particularly, AWS’ cloud environments. In fact, none of the pretenders to the infrastructure crown that really use a Zen-based platform, have achieved anything spectacular in Europe. It seems to be more the traditional vendors, like VMware and others, that are gaining confidence in the public and government spaces as well as larger businesses.
GC: In the business level conversations that you’re having with companies or vendors, what’s the strongest justification for moving to the cloud?
AP: I would have to continue to say that the strongest justification, when all is said and done, continues to be driven primarily from an economic or a financial perspective. Even a move from CAPEX to operational expenditure (OPEX) has always been the value proposition of any outsourced IT services or managed services. Instead of running around asking for more capital to build a data center or even if it’s just the need for new servers or software, cloud continues to be a big draw. And even if it’s not the capital costs, there tends to be a lot of emphasis around reducing operational costs. This means if I can hire one less person and have a really commoditized service like exchange services, so my email service is run by somebody who gets economies of scale, I’m going to do that because paying $2,000 a month to manage process and manage my exchange, is a lot less expensive than the IT Admin guy who would be running it otherwise.
And that’s just in its simplest form. Expand that out to the rest of the applications, CRM and any other business applications, and it’s the same premise. Now, for cloud computing, which is meant to be far more flexible than just traditional managed or dedicated hosting, the real advantage there moves slightly away from just the financial benefits and closer towards the change from cost efficiency to operational efficiency. With cloud, people can do things so much faster and with much better tools than they ever did in the past.
And time is money. Just the fact that you can now control to this level of detail and nuance is amazing. And if you take a little side step to ScienceLogic, it’s the same thing: some of the tools from yesteryear, even we were not able to make that leap into this cloud computing world because you could get tools to manage your servers and systems but they tended not to be multi-tenant tools. In other words, you have to go in and log into each server, one at a time, and use the tools to configure something or to check something or monitor something. Now, those days are gone. You don’t need to do that anymore in the era of cloud computing where things move fast. People expect things to be set up far quicker, so you don’t want to waste time re-configuring everything.
GC: How do you view cloud moving into the enterprise?
AP: Five years ago when we spoke to the big service providers, a lot of them were not very eager. AT&T was one of them and that’s why they were kind of late to the game when it came to cloud computing services. In the UK, Colts were kind of late to the game, too. They were very hesitant to use the word “cloud” for that very reason: the enterprises were rolling their eyes at cloud computing, saying it was the “same old, same old,” that nothing had changed, and, if anything, it had a negative connotation because it conjured up images of a very soft, not very hardened or hardcore security process; generally not a very robust type of service.
That image has changed now. Everyone has recognized the fact that cloud is not just a fleeting moment. Rather, everyone’s using it: everyone in the consumer space; everyone in the mobile space; everyone in the infrastructure space and so on. It’s become far more serious and far more acceptable, so much so that even the software vendors who enable cloud computing are focused on cloud enablement. Cloud is not just here to stay, but it is here to grow. The enterprise sector has been very slow to jump on this bandwagon for a number of reasons, but they’re finally there. However, their flavor of cloud computing is really being driven by private cloud.
It’s no secret that 85% of IT infrastructure continues to be in-house. The next generation upgrade in hardware, software and management tools all represent what enterprises consider private cloud. But it is an evolution. When I look at some of the analysts’ reports like InfoPro and Gartner, there’s an acknowledgment by a lot of the enterprises that this is going to be an evolution; they’re going to need to pony up the money to upgrade their internal cloud environment, their private cloud, and then expand that same footprint with the same technologies into an external data center so it becomes a hybrid cloud. They’re going to have a completely private in-house, on-premise cloud environment and then a virtual hosted private cloud environment at a third party. But what’s important is that the third party has the same tools and the same technologies that the enterprise has decided upon in-house.
So for the enterprise – and this is a big difference – whereas small business and mid-sized business are dictated to by the service providers in terms of what technology is acceptable, big business seems to be dictating to their service providers what kind of technology they want and what kind of tools they want to be using.
The evolution is coming, but it is different in that it’s starting in-house and the enterprises are dictating what they want to see. They are being less influenced by the “circle of trusted advisors” provider relationships.
GC: What are the chief internal and external barriers to cloud adoption for the enterprise?
AP: The same thing that’s stopping it is the same thing that’s promoting it – the IT department. But it’s the rogue IT department that’s really leading the way more than anything else. There are other terms to use other than “rogue” but it’s definitely still being led by independent departments and organizations, so it continues to be more of a tactical measure rather than a holistic strategic measure taken by the enterprises.
I feel like the enterprises are where the mid-sized companies were five years ago; they really are that far behind. Cloud represents a cultural change. It’s a change of acceptance and work assurance: “Am I still going to be useful here if I don’t accept cloud?” So, I think job safety is the number one hindrance to the individual IT guys, which dovetails into issues related to cultural change across the board. Everyone is going to have to accept that there is going to be a new way of consuming IT and that it’s not necessarily going to be under internal controls, or that the budget may not necessarily reside with individual departments anymore.
What cloud represents is the opportunity to really re-use the same infrastructure at various times of the day, for various purposes. So the finance department can use it once a month for accounting software for their invoices. Another group can use it at another time for another purpose, and that trend will continue. That kind of flexibility is really what helps these departments, but it’s that acceptance which is extremely hard. Also, to that end, if there is no executive sponsorship or buy-in at the top, then it’s all hindrance.
The next biggest set of hindrances is really around compliance and governance. Governance, in this case, also meaning internal governance and I’ll give you an example: there are large hospitals and medical centers that simply will not allow any kind of application or development or processing to take place off their premise. Even if they are unsure that it breaks any rules or any regulations, they won’t risk it. I don’t believe that security is a hindrance, it’s merely a hurdle that needs to be overcome, and everybody needs to answer it.
dIronically, another secondary hindrance is cost. Even though it is the big motivator, depending on who the enterprise ends up speaking to, that service provider can actually also turn them off. For example, let’s say you go in with your systems admins and your data center people, and you’re looking for a way to expand your data center without having to spend a huge chunk of money. Any other outsourced data center service, even if it is an infrastructure-as-a-service or the co‑location of a data center, is going to be very convenient to you. But if you’re looking for an application-level service and you’re looking for software-as-a-service opportunities, it can be very useful for you if you’re accepting of a cookie-cutter solution. But as soon as you want something customized and bespoke, it can get very expensive. And so can infrastructures-as-a-service. AWS has a lot of detractors because it can get very expensive and you can lose control of your costs very easily too. So that issue in particular becomes a barrier in terms of adoption. People get very queasy about losing control of their costs. So it’s ironic; it can be a driver, a motivator for some companies, and a hindrance for others.
By Jake Gardner