A new year is often a time for reflection on the past and pondering the future. 2010 was certainly a momentous year for cloud computing. An explosion of tools for creating clouds, a global investment rush by service providers, a Federal “cloud first” policy, and more. But in the words of that famous Bachman Turner Overdrive song — “You ain’t seen nothin’ yet!”
In fact, I’d suggest that in terms of technological evolution, we’re really just in the Bronze Age of cloud. I have no doubt that at some point in the not too distant future, today’s cloud services will look as quaint as an historical village with no electricity or running water. The Wired article on AI this month is part of the inspiration for what comes next. After all, if a computer can drive a car with no human intervention, why can’t it run a data center?
Consider this vision of a future cloud data center.
The third of four planned 5 million square foot data centers quietly hums to life. In the control center, banks of monitors show data on everything from number of running cores, to network traffic to hotspots of power consumption. Over 100,000 ambient temperature and humidity sensors keep track of the environmental conditions, while three cooling towers vent excess heat generated by the massively dense computing and storage farm.
The hardware, made to exacting specifications and supplied by multiple vendors, uses liquid coolant instead of fans – making this one of the quietest and most energy-efficient data centers on the planet. The 500U racks reach 75 feet up into the cavernous space, though the ceiling is yet another 50 feet higher where the massive turbines draw cold air up through the floors. Temperature is relatively steady as you go up the racks due to innovative ductwork that vents cold air every 5 feet as you climb.
Advanced robots wirelessly monitor the 10GBps data stream put off by all of the sensors, using their accumulated “knowledge and experience” to swap out servers and storage arrays before they fail. Specially designed connector systems enable individual pieces or even blocks of hardware to be snapped in and out like so many Lego blocks – no cabling required. All data moves on a fiber backbone at multiple terabytes per second.
On the data center floor, there are no humans. The PDUs, cooling systems and even the robots themselves are maintained by robots – or shipped out of the data center into an advanced repair facility when needed. In fact, the control center is empty too – the computers are running the data center. The only people here are in the shipping bay, in-boarding the new equipment and shipping out the old and broken, and then only when needed. Most of these work for the shippers themselves. The data center has no full-time employees. Even security and access control for the very few people allowed on the floor for emergencies is managed by computers attached to iris and handprint scanners.
The positioning and placement of storage and compute resources makes no sense to the human eye. In fact, it is sometimes rearranged by the robots based on changing demands placed on the data center – or changes that are predicted based on past computing needs. Often this is based on private computing needs of the large corporate and government clients who want (and will pay for) increased isolation and security. The bottom line – this is optimized far beyond what a logical human would achieve.
Tens of millions of cores, hundreds of exabytes of data, no admins. Sweet.
The software automation is no less impressive. Computing workloads and data are constantly optimized by the AI-based predictive modeling and management systems. Data and computing tasks are both considered to be portable – one moving to the other when needed. Where large data is required, the compute tasks are moved to be closer to the data. When only a small amount of data is needed, it will often make the trip to the compute server. Of course, latency requirements also play a part. A lot of the data in the cloud is maintained in memory — automatically based on demand patterns.
The security AI is in a constant and all-out running battle with the bots, worms and viruses targeting the data center. All server images are built with agents and monitoring tools to track anomalies and attack patterns that are constantly updated. Customers can subscribe to various security services and the image management system automatically checks for compliance. Most servers are randomly re-imaged throughout the day based on the assumption that the malware will eventually find a way to get in.
Everything is virtualized – servers, storage, networking, data, databases, application platforms, middleware and more. And it’s all as a service, with unlimited scale-out (and scale-in) of all components. Developers write code, but don’t install or manage most application infrastructure and middleware components. It’s all there and it all just works.
Component-level failure is assumed and has no impact on running applications. Over time, as the AI learns, reliability of the software infrastructure underlying any application exceeds 99.999999%.
Everything is controllable through APIs, of course. And those APIs are all standards-based so tools and applications are portable among clouds and between internal data centers and external clouds.
All application code and data is geographically dispersed so even the failure of this mega data center has a minimal impact on applications. Perhaps a short hiccup is experienced, but it lasts only seconds before the applications and data pick up and keep on running.
Speaking of applications, this cloud data center hosts thousands of SaaS solutions for everything from ERP, CRM, e-commerce, analytics, business productivity and more. Horizontal and vertical applications too. All exposed through Web services APIs so new applications – mashups – can be created that combine them and the data in interesting new use cases. The barriers between IaaS, PaaS and SaaS are blurred and operationally barely exist at all.
All of this is delivered at a fraction of the cost of today’s IT model.
Large data center providers using today’s automation methods and processes are uncompetitive. Many are on the verge of going out of business and others are merging in order to survive. A few are going into higher-level offerings – creating custom solutions and services.
The average enterprise data center budget is 1/10th of what it used to be. Only the applications that are too expensive to move or otherwise lack suitability for cloud deployment are still in-house managed by an ever-dwindling pool of IT operations specialists (everybody else has been retrained in cloud governance and management, or found other careers to pursue). Everything else is either a SaaS app or otherwise cloud-hosted.
Special-purpose clouds within clouds are easily created on the fly, and just as easily destroyed when no longer needed.
The future of the cloud data center is AI-managed, highly optimized, and incredibly powerful at a scale never before imagined. The demand for computing power and storage continues to grow at ever increasing rates. Pretty soon, the data center described above will be considered commonplace, with scores or even hundreds of them sprinkled around the globe.
This is the future – will you be ready?
Follow me on twitter for more cloud conversations: http://twitter.com/cloudbzz
Notice: This article was originally posted at http://CloudBzz.com by John Treadway.
(c) CloudBzz / John Treadway
OK, so let’s say this happens. I think what’s more interesting are the implications for incumbent IT vendors. What effect do you expect this will have on them and who will be the emerging new “winners? Your emphasis on AI (now called machine learning) suggests the value moves from hardware vendors to smart, active, automated management software. What’s your vision here?
Cloud computing is already disrupting IT vendors. Look at what Cloudscaling did at KT with open source, commodity hardware, and a few tools from newer vendors (Cloud.com and enStratus among others). nnThere’s no technical reason why HW vendors can’t add this level of intelligence to their systems. However, that’s not been the trend (UCS might be an exception). nnThe new technology will come from government projects, mega data center providers (e.g. Google) and others who have the $ and time to get it working — and the economic incentive to do so. BMC and others will try but their legacy investments and entrenched internal interests will make this an uphill battle. They can do it, but only by a willingness to cannibalize their own businesses.nnI expect new hardware models will also appear – challenging the standards rack and blade-mounted systems. Way too much redundancy, too many connection points, not modular enough (even blades), and too power-hungry. In fact, the biggest challenge to this vision is not technology – it’s power. Lower power chips, SSDs, and fast distributed bus technologies might replace today’s ear-damage-inducing mega-farms.nnHey, even my MacBook Air has two cores, 4GB of RAM (could easily be 64GB like my iPad), and the fan rarely runs. It runs VMware nicely and can probably handle 8-10 small Linux VMs quite easily. Get rid of the screen beautifully engineered case, internal battery, wifi, bluetooth, USB, DVI, etc. and this would be a fair cloud server for less than $500. Okay, that might be a stretch, but it’s not a big one. This sucker’s fast, quiet and cool to the touch. You could put two such “blades” in the space of a current drive bay for a storage array. No case – just the board open to the circulating air sitting on a water-cooled heat sink. Do it with Snapdragons or ARMs at ~10w per server (vs. 400W+ today). nnI predict that in 5-7 years a new crop of adaptive, machine learning (AI) cloud management technologies will be running rings around today’s “enterprise” orchestration and provisioning tools with their hand-scripted runbooks and CMDBs. The win would be enormous, and the winners will be rich.
This doesn’t sound so different from the way we did it at Exodus Communications 14 years ago, except we had people not robots moving boxes around. I love the vision and most of it is quite accurate, with one big exception. The idea that a server is a discrete unit that people will pay for just doesn’t hold water over time. From HP’s original Computon to Amazon’s ECU, we’ve been evolving transaction or processor-based pricing. Software architects should be able to effectively ignore the concept of a server, as a data center IS a big SMP server (as I described here!) http://www.nytimes.com/external/gigaom/2009/06/14/14gigaom-what-intel-can-teach-google-about-the-cloud-92958.htmlnnThe other thing that’s obsolete is the operating system. It all started in about 2003 when VMware ESX stopped installing on top of a host OS, instead becoming the host OS for data center operations. nnTake everything you said in this visionary piece, and assume that coders write apps that run in containers and pay per use…or per MIPS in some cases. A good cloud provider will (out of economic necessity) take poorly written code and optimize it to run most efficiently. An early example is Google’s Website Optimizer that rewrites bad HTML to be faster. nnYour comments on security are accurate too – there will need to be a variety of cloud security vendors because no cloud infrastructure vendor is going to focus exclusively on security. Game theory says they will focus on enough security to make the most profits, but not as much security as customers want. More on that here. http://cloudsecurity.trendmicro.com/conflict-of-interest-leads-to-big-malware-attack/nnCongratulations on a great post!
Dave – thanks and regarding physical servers I completely agree. There is an exception – when a customer wants a discrete set of physical infrastructure for a private cloud. The developers should not know or care about the underlying boxes, in that we agree fully. Interesting post on game theory. At the end of the day that’s how all decisions are made — weighing costs and risks and hopefully getting the balance of the two more right than wrong. nnAs for ESX being an operating system, that’s still not really true. But it can be at some level. The issue is that developers still like to optimize things (e.g. build my variant of linux with this compiled in and that compiled out…). You don’t want that happening against bare metal so the hypervisor will need to host operating systems to host apps for a while longer.nnJohn
“I predict that in 5-7 years a new crop of adaptive, machine learning (AI) cloud management technologies will be running rings around today’s “enterprise” orchestration and provisioning tools with their hand-scripted runbooks and CMDBs.” – Exactly! This is a post in itself. Looking forward to it 🙂
Data Center Rack Security Solutions manufactured by Netrack offers best quality Monitoring option for the Server personnel and Includes finger print access in India