22 November 2013 Posted by Paul Burns
In mid-November I had the pleasure of attending re:Invent, the conference for members of the AWS community including developers, IT leaders and partners. I spent one full day participating in the industry analyst track with members of the AWS management team as well as some AWS customers. I spent another day meeting with a variety of AWS partners. re:Invent is THE conference for AWS and the AWS ecosystem. It is also a great conference for cloud computing in general.
One of the themes most emphasized by Amazon in the conference was, “AWS is ready for the enterprise.” Similarly, one of the most discussed topics since the conference has been, “Is AWS ready for the enterprise?”
Enterprise Ready?
While it is clear to me that this is not a simple yes or no question, that hasn’t stopped a few people from answering it that way. Strictly affirmative answers have included:
- Enterprise IT organizations are using AWS; so it must be enterprise ready.
- Enterprise “A” is running production applications – or even a mission-critical application – on AWS, thus it is enterprise ready.
- Enterprise “B” eliminated its data center and replaced all in-house IT with AWS. Therefore AWS is enterprise ready.
Strictly negative answers have included:
- AWS has suffered from extended outages; so it must not be enterprise ready.
- Netflix – the most common reference customer for AWS adoption – only runs one application on AWS and it was built for AWS. Enterprise IT organizations have hundreds or even thousands of applications, which were not built for AWS. Thus AWS is not enterprise ready.
- Some applications must be re-architected to run safely on AWS. Therefore AWS is not enterprise ready.
There are of course many other variations to these basic arguments. However, these black and white answers do very little to answer the enterprise-readiness question.
The Bigger Picture
In order to answer – or even just address – the question of enterprise readiness with an appropriate level of nuance, it is helpful to keep the changing state of the IT industry in mind, along with some of the stronger competitive forces.
Consider this image of a partially constructed bridge:
This image serves as a metaphor for the emergence and growth of cloud computing in general – not just AWS. While not explicitly shown, I’ll also add the claim that the direction of the future for IT is “generally” moving from left to right.
I’m not saying public cloud or even public plus private cloud will completely replace enterprise IT as it currently exists. While many people do indeed say that, I believe – and history seems to show – that enterprise IT tends to add new technologies, processes and business models rather than replace them. Additionally, any time horizon I can imagine where cloud could “possibly” replace the vast majority of enterprise IT, is too long to reliably predict what IT will look like, with or without cloud computing.
I’m really suggesting that the current direction of IT is toward some of the attributes that cloud computing supports particularly well, including elasticity, on-demand, scalable, automated, API-enabled infrastructure consumption. Could some of this direction change or some of the related needs be met in other ways? Certainly. However, I believe the agility, economics and efficiency of cloud computing makes it highly likely that cloud computing – in the forms of public, private and hybrid IaaS – will increasingly become an important building block for enterprise IT.
What else should this image convey? Here are some general characterizations, keeping in mind that conflicting examples can likely be found for each point:
Enterprise IT:
- Proven and working, with problems of its own – sometimes including high costs and lack of flexibility
- Refined over decades to support existing use cases
- Plenty of innovation in many areas of hardware and software, though less fundamental or disruptive than on the cloud side
Cloud Computing:
- Proven and working for many use cases – not proven for all enterprise IT use cases, but for many of them
- Has new attributes (e.g. elasticity, speed of change…) that enable IT to do some things differently or do things they previously could not do
- Represents more disruptive innovation, impacting industry competitive dynamics and delivering new types of value, sometimes in a format difficult for enterprise IT to initially adopt
There are of course many other ways to slice and dice these two approaches to computing and many other attributes to associate with each. The key idea is that a disruptive gap currently exists between traditional enterprise IT and cloud computing. Enterprise IT is working, but certainly has room for improvement. At the same time, there are unique and powerful benefits found on the right side of the gap. However, it is not always easy for enterprise IT to utilize them and cloud computing does not currently offer a full and complete replacement to enterprise IT.
Assessing Enterprise Readiness
Some still seem to want to say that AWS is already a replacement solution for enterprise IT. Others seem to want to say that AWS is not ready to support enterprise IT and never will be ready. The partially constructed bridge metaphor offers a fairer, more accurate way to look at things.
Left side of the bridge
For now, the worlds of enterprise IT and cloud computing remain divided to some degree. With this gap in place, the incumbents on the left (e.g. traditional IT vendors and service providers) still have a massive lead in serving enterprise IT organizations. This is where the vast majority of infrastructure and applications currently reside. Most of the current enterprise IT spending also continues to happen here, for now.
The entire set of hardware, software and services offered by vendors and non-cloud service providers on the left side of the bridge is currently better at meeting the needs of enterprise IT than the entire set of cloud service providers on the right hand side. That is the current situation. If you don’t like it, get over it. ☺
Right side of the bridge
On the cloud computing side of the bridge, in the infrastructure as a service (IaaS) segment, AWS is the clear market leader in terms of revenue and, arguably, several other dimensions such as breadth of services. In fact, AWS has a very strong and defendable position on the right hand side of the gap. I have separately identified a dozen or so ways that AWS is uniquely maintaining and building its lead in IaaS+. There are also several areas where other providers have stronger offerings, particularly within compute services. All that is a separate topic.
The entire set of IaaS providers on the right hand side of the bridge is collectively growing at a faster rate than the enterprise IT industry on the left. And the services offered on the right hand side of the bridge are far better at serving certain needs such as massive scale, rapidly fluctuating demand and agile IT. That is the current situation. If you don’t like it, get over it. ☺
Building the Bridge
Interestingly, industry participants on both sides of the bridge have a stake in building it. And not building it. And building it differently than the other side prefers. This is where things gets complicated.
Many of those on the right hand side prefer a minimalist bridge. “Lift and shift, baby! All you need to do is migrate your virtual machines from on-premises servers to our cloud and you are set!” Or, how about new models for IT. “Come on over to the cloud side for things like DevOps. There is no way to do DevOps on that tired, old, fixed infrastructure of yours.”
Things are every bit as one sided on the left hand side of the bridge. “The cloud is unreliable! Keep your applications on premises because they won’t run reliably on a cloud.” Or, how about security. “Keep it secret. Keep it safe. If you want your data safe, you must keep it on premises. The NSA can scrape cloud data but not enterprise data.”
Bridges will be Built
It is an ugly process, but bridges will be built. And have been built. And will also, in some cases, be blocked.
But, for now, it is not fair to give a black and white answer – either affirmative or negative – to AWS, or IaaS in general, being enterprise ready. It depends on the application. It depends on the data. It depends on applicable governance standards. It depends on the availability of management tools that can see deep enough inside public cloud networks to enable troubleshooting of anomalies that impact user experience. It depends on the consumption patterns of IT resources and the economic attributes and behaviros of the underlying infrastructure. It depends on the culture of the enterprise IT organization in question. It depends on the strategy and requirements of the business organizations which IT supports.
Of course many of these “it depends” items represent the metaphorical bridge(s) that are being built. This means that the answers are also changing over time, as the bridges are built. Cloud computing is becoming more enterprise ready ever day.
Enterprise readiness in the IaaS space also depends on the service provider itself. AWS is pursuing developers as customers and providing many services above and beyond what defined the IaaS layer just a few years ago. This often makes AWS the best choice for rapidly testing new ideas or building new, cloud aware applications. Still, other IaaS players are pursuing an IaaS 2.0 model, delivering higher performance, more consistent performance, lower prices and/or offering other capabilities that make their clouds more attractive than AWS for many enterprise applications.
Neovise completed research with hundreds of cloud adopters in 2013. The results showed strongly that enterprise IT organizations are choosing mulitple clouds. This includes mulitple public, private and hybrid clouds. These organizations are matching applications to the cloud that best meets their needs. They are also keeping applications within existing on-premises environments when that makes the most sense. And it often does. This alone shows that AWS is not somehow the single answer for enterprise IT in the IaaS space.
Enterprise IT organizations are absolutely using AWS – and other public IaaS clouds – and running production applications. Some are successfully running what could be considered mission-critical applications that were not initially designed to run in an IaaS environment. I have even interviewed some small and mid-sized businesses that have eliminated and/or are are working toward eliminating their use of their own private data centers to go “all in” with AWS (These companies identifed themselves as “enterprise” but were on the order of $1M to $10M in annual revenue). Some $B+ / year revenue enterprises are also rumored to be moving large percentages of their workloads to AWS and/or other public clouds. This is all great progress for enterprise readiness in cloud computing. It just doesn’t mean that it is somehow time to move all of enterprise IT to cloud computing. Far from it.
Conclusions
As illustrated by the partially built bridge, a disruptive gap currently exists between traditional enterprise IT and cloud computing. In some cases the best place for enterprise IT is within traditional on-premises environments. In other cases the best place for enterprise IT is within a public IaaS cloud. When a public IaaS cloud is the right answer, AWS may or may not be the best choice for the service provider. AWS is by many measures the dominant IaaS provider – AND, has a highly defendable position, which I can discuss separately. It just hasn’t magically made itself the best answer for all of enterprise IT – and there are competing IaaS providers with better offerings than AWS for various enterprise IT scenarios.
There are many good strategies for cloud adopters to follow. Whether you live strictly on the left or right side of the bridge, or you are participating in building a bridge, or you are “all in” on the right side of the bridge – cloud computing is still in the process of maturing. And it will continue disrupting enterprise IT as it matures. Enterprise IT must keep an eye on the bridges being built in order to keep making optimal deployment decisions.
There are also many good strategies for vendors and service providers within the IT industry to follow. AWS will continue to be a large part of cloud computing and a growing part of enterprise IT. Traditional enterprise IT vendors and service providers will also continue building bridges – and taking advantage of bridges built by others – to strengthen their positions across the gap in the world of cloud computing. Of course some won’t make it.
AWS competitors that live strictly within the cloud computing world certainly have their hands full, but there are also many strategies for them to continue competing and growing. That is yet another great topic for another time.
]]>
Recent blog posts
- Revisiting the Benefits of SaaS
- Compliance Drives Buying Decisions in the Cloud
- Changing Competitive Dynamics for IaaS Providers
- Good News for Cloud Job Seekers in 2015
- Colocation: The New Center of Hybrid IT
- The Problem with Forced Cloud Reboots
- OnApp Federation Grows Stronger with SolusVM Acquisition
- HP to Acquire Eucalyptus: Questions, Comments and Concerns
- Is It Really ‘Game Over’ for Most IaaS Providers?
- The Big Cloud Winners Are…Not Just the Hyperscale Players