Archive

Archive for the ‘workload’ Category

Why we about to move from cloud computing to industrial computing?

April 5, 2010 7 comments

I spent the other week at a new conference called Cloud Connect. Being able to spend four days emerged in an industry discussion about cloud computing really allows you to step back and think about where we are with this emerging industry. While it would be possible to write endlessly about all the meeting and conversations I had, you probably wouldn’t have enough time to read all that. So, I’ll spare you and give you the top four things I learned at Cloud Connect. I recommend that you also take a look at Brenda Michelson’s blogs from the event for a lot more detail. I would also refer you to Joe McKendrick’s blog from the event.

1. Customers are still figuring out what Cloud Computing is all about.  For those of us who spend way too many hours on the topic of cloud computing, it is easy to make the assumption that everyone knows what it is all about.  The reality is that most customers do not understand what cloud computing is.  Marcia Kaufman and I conducted a full day workshop called Introduction to Cloud. The more than 60 people who dedicated a full day to a discussion of all aspects of the cloud made it clear to us that they are still figuring out the difference between infrastructure as a service and platform as a service. They are still trying to understand the issues around security and what cloud computing will mean to their jobs.

2. There is a parallel universe out there among people who have been living and breathing cloud computing for the last few years. In their view the questions are very different. The big issues discussed among the well-connected were focused on a few key issues: is there such a thing as a private cloud?; Is Software as a Service really cloud computing? Will we ever have a true segmentation of the cloud computing market?

3. From the vantage point of the market, it is becoming clear that we are about to enter one of those transitional times in this important evolution of computing. Cloud Connect reminded me a lot of the early days of the commercial Unix market. When I attended my first Unix conference in the mid-1980s it was a different experience than going to a conference like Comdex. It was small. I could go and have a conversation with every vendor exhibiting. I had great meetings with true innovators. There was a spirit of change and innovation in the halls. I had the same feeling about the Cloud Connect conference. There were a small number of exhibitors. The key innovators driving the future of the market were there to discuss and debate the future. There was electricity in the air.

4. I also anticipate a change in the direction of cloud computing now that it is about to pass that tipping point. I am a student of history so I look for patterns. When Unix reached the stage where the giants woke up and started seeing huge opportunity, they jumped in with a vengeance. The great but small Unix technology companies were either acquired, got big or went out of business. I think that we are on the cusp of the same situation with cloud computing. IBM, HP, Microsoft, and a vast array of others have seen the future and it is the cloud. This will mean that emerging companies with great technology will have to be both really luck and really smart.

The bottom line is that Cloud Connect represented a seminal moment in cloud computing. There is plenty of fear among customers who are trying to figure out what it will mean to their own data centers. What will the organizational structure of the future look like? They don’t know and they are afraid. The innovative companies are looking at the coming armies of large vendors and are wondering how to keep their differentiation so that they can become the next Google rather than the next company whose name we can’t remember. There was much debate about two important issues: cloud standards and private clouds. Are these issues related? Of course. Standards always become an issue when there is a power grab in a market. If a Google, Microsoft, Amazon, IBM, or an Oracle is able to set the terms for cloud computing, market control can shift over night. Will standard interfaces be able to save the customer? And how about private clouds? Are they real? My observation and contention is that yes, private clouds are real. If you deploy the same automation, provisioning software, and workload management inside a company rather than inside a public cloud it is still a cloud. Ironically, the debate over the private cloud is also about power and position in the market, not about ideology. If a company like Google, Amazon, or name whichever company is your favorite flavor… is able to debunk the private cloud — guess who gets all the money? If you are a large company where IT and the data center is core to how you conduct business — you can and should have a private cloud that you control and manage.

So, after taking a step back I believe that we are witnessing the next generation of computing — the industrialization of computing. It might not be as much fun as the wild west that we are in the midst of right now but it is coming and should be here before we realize that it has happened.

Why hardware still matters– at least for a couple of years

February 9, 2010 3 comments

It is easy to assume that with the excitement around cloud computing would put a damper on the hardware market. But I have news for you. I am predicting that over the next few years hardware will be front and center.  Why would I make such a wild prediction. Here are my three reasons.

1. Hardware is front and center in almost all aspects of the computer industry. It is no wonder that Oracle wants to become a hardware company. Hardware is tangible. It’s revenue hits the bottom line right away. Hardware can envelop software and keep customers pinned down for many, many years. New generation platforms in the form of hardware appliances are a convenient delivery platform that helps the sales cycle. It is no wonder that Oracle wants a hardware platform. It completes the equation and allows Oracle to position itself as a fully integrated computing company. Likewise, IBM and HP are focused on building up their war chest full of strong hardware platforms. If you believe that customers want to deal with one large brand..or two, then the winners want to control the entire computing ecosystem.

2. The cloud looms. Companies like Amazon.com and Google do not buy hardware from the big iron providers and never will. For economic reasons, these companies go directly to component providers and purchase custom designed chips, board, etc. This approach means that for a very low price, these cloud providers can reduce their power consumption by making sure that the components are optimize for massively scaled clouds.  These cloud vendors are focused on undercutting the opportunity and power of the big systems providers. Therefore, cloud providers care a lot about hardware — it is through optimization of the hardware that they can threaten the power equilibrium in the computer market.

3. The clash between cloud and on premise environments. It is clear that the computer marketplace is at a transition point. The cloud vendors are betting that they can get the costs based on optimization of everything so low that they win. The large Systems vendors are betting that their sophisticated systems combining hardware, software, and service will win because of their ability to better protect the integrity of the customer’s business. These vendors will all provide their own version of the public and private cloud to ensure that they maintain power.

So, in my view there will be an incredible focus on hardware over the next two years. This will actually be good for customers because the level of sophistication, cost/performance metrics will be impressive. This hardware renaissance will not last. In the long run, hardware will be commoditized. The end game will be interesting because of the cloud. It will not a zero sum game. No, the data center doesn’t go away. But the difference is that purpose built hardware will be optimized for workloads to support the massively scaled environments that will be the heart of the future of computing. And then, it will be all about the software, the data, and the integration.

3.

Tectonic shifts: HP Plus 3Com versus Cisco Plus EMC

November 18, 2009 4 comments

Just when it looked clear where the markets were lining up around data center automation and cloud computing, things change. I guess that is what makes this industry so very interesting.  The proposed acquisition by HP of 3Com is a direct challenge to Cisco’s network management franchise. However, the implications of this move go further than what meets the eye.  It also pits HP in a direct path against EMC with its Cisco partnership. And to make things even more interesting, it also puts these two companies in a competitive three way race against IBM and its cloud/data center automation strategy. And of course, it doesn’t stop there. A myriad of emerging companies like Google and Amazon want a larger share of the enterprise market for cloud services. Companies like Unisys and CSC that has focused on the outsourced secure data centers are getting into the act.

I don’t think that we will see a single winner — no matter what any one of these companies will tell you.  The winners in this market shift will be those companies can build a compelling platform and a compelling value proposition for a partner ecosystem.  The truth about the cloud is that it is not simply a network or a data center. It is a new way of providing services of all sorts that can support changing customer workloads in a secure and predictable manner.

In light of this, what does this say for HP’s plans to acquire 3Com? If we assume that the network infrastructure is a key component of an emerging cloud and data center strategy, HP is making a calculated risk in acquiring more assets in this market.  The company that has found that its ProCurve networking division has begun gaining traction. HP ProCurve Networking is the networking division of HP.  The division includes network switches, wireless access points, WAN routers, and Access Control servers and software.   ProCurve competes directly with Cisco in the networking switch market. When HP had a tight partnership with Cisco, the company de-emphasized the networking. However, once Cisco started to move into the server market, the handcuffs came off. The 3Com acquisition takes the competitive play to a new level. 3Com has a variety of good pieces of technology that HP could leverage within ProCurve. Even more significantly, it picks up a strong security product called TippingPoint, a 3Com acquisition. TippingPoint fills a critical hole in HP’s security offering. TippingPoint, offers network security offerings including intrusion prevention and a product that inspects network packets.  The former 3Com subsidiary has also established a database of security threats based a network of external researchers.

But I think that one of the most important reasons that HP bought 3Com is its strong relationships in the Chinese market. In fiscal year 2008 half of 3Com’s revenue came from its H3C joint venture with Chinese vendor, Huawei Technology. Therefore, it is not surprising that HP would have paid a premium to gain a foothold in this lucrative market. If HP is smart, it will do a good job leveraging the many software assets to build out both its networking assets as well as beefing up its software organization. In reality, HP is much more comfortable in the hardware market. Therefore, adding networking as a core competency makes sense. It will also bolster its position as a player in the high end data center market and in the private cloud space.

Cisco, on the other hand, is coming from the network and moving agressively into the cloud and the data center market.  The company has purchased a position with VMWare and has established a tight partnership with EMC as a go to market strategy.  For Cisco, it gives the company credibility and access to customers outside of its traditional markets. For EMC, the Cisco relationship strengthens its networking play.  But an even bigger value for the relationship is to present a bigger footprint to customers as they move to take on HP, IBM, and the assortment of other players who all want to win.  The Cisco/EMC/VMware play is to focus on the private cloud.  In their view a private cloud is very similar to a private, preconfigured data center.  It can be a compelling value proposition to a customer that needs a data center fast without having to deal with a lot of moving parts.  The real question from a cloud computing perspective is the key question: is this really a cloud?

It was inevitable that this quiet market dominated by Google and Amazon would heat up as the cloud becomes a real market force.  But I don’t expect that HP or Cisco/EMC will have a free run. They are being joined by IBM and Microsoft — among others. The impact could be better options for customers and prices that invariably will fall. The key to success for all of these players will be how well they manage what will be an increasingly heterogeneous, federated, and highly distributed hardware and software world. Management comes in many flavors: management of these highly distributed services and management of the workloads.

Why all workloads don’t belong in the cloud

November 2, 2009 5 comments

I had an interesting conversation with a CIO the other day about cloud computing. He had a simple question: I have an relatively old application and I want to move it to the cloud. How do I do that? I suspect that we will see a flurry of activity over the coming year where this question will be asked a lot.  And why not — the cloud is the rage and who wouldn’t want to demonstrate that with the cloud all problems are solved.  So, what was my answer to this CIO? Basically, I told him that all workloads do not belong in the cloud. It is not because this technically can’t be done. It can. It is quite possible to encapsulate an existing application and place it into a cloud environment so that new resources can be self-provisioned, etc. But, in reality, you have to look at this issue from an efficiency and an economic perspective.

ROI

Cloud computing gains an economic edge over a traditional data center when it supports a relatively small simple workload for a huge number of customers. For example, a singular workload like email or a payment service can be fairly optimized at all levels — the operating system, middleware, and the hardware can all be customized and tuned to support the workload. The economics favor this type of workload that support large numbers of customers. The same cannot be said for the poor aging Cobol application that is used by 10 people within an organization. While there might be incremental management productivity benefits, the cost/benefit analysis simply doesn’t work.

So, the answer is pretty simple. You just can’t throw every workload into the cloud. It is not a panacea for all IT problems.  Organizations that are trying to figure out what to do with these pesky old workloads need to look at three options:

1. Decide if that workload is still supporting business objectives in a cost effective manner. If it does the job, leave it alone.

2. That old workload might be better supported by traditional outsourcing. Let someone else keep the application alive while you move into more mission critical tasks.

3. Think about rebuilding that old workload — either by encapsulating key elements and placing them within a modular flexible environment. You might even discover that there are components that are actually useful across the organization. When you discover that sharing components across divisions/department is a productive and pragmatic approach, you might be ready to move those workloads into the cloud.

cloud_box