Today (November 3, 2009), VMware, Cisco, and EMC announced an alliance called VCE. The coalition packages up infrastructure based primarily on VMware Inc. (NYSE: VMW) software (vSphere); Cisco Systems Inc. (Nasdaq: CSCO) compute and network infrastructure (UCS and Nexus switches); and EMC Corp. (NYSE: EMC) storage.
This is an important milestone because it highlights the state of cloud computing, where it fits and where it doesn’t -- a subject I’ve been talking about to IT practitioners and technologists since I last examined cloud computing in detail about six months ago.
I want to share with you the results of some of my research and provide a checkpoint on cloud adoption and competition, with an eye on today’s announcement.
Here’s the high-level summary: There are two clouds -- one that is new and different and one that’s all about virtualizing the existing data center.
The “new cloud” has the following characteristics:
The new cloud is highly fragmented and growing like crazy, with an ecosystem that is spawning new companies and technologies like Nirvanix Inc. and Heroku.
The new cloud is highly virtualized, homogeneous, object-based, RESTful, multi-tenant, pay-as-you-go.
The new cloud is always about connecting to computing resources through the Internet and supporting hundreds of thousands -- or millions -- of connected organizations.
Compare these features to simply virtualizing an existing infrastructure:
The virtualized data center is usually about organizing resources behind a firewall for a single customer.
The virtual data center is all about consolidating big companies, integrating stacks, re-aligning ecosystems, and trying to hold onto market share or grow by stealing someone else’s share -- essentially a zero-sum game.
The virtual data center is slightly virtualized, heterogeneous, file- and block-based, RESTless, and anything but pay-as-you-go, at least today.
In squinting through the hype, you start to see that the new cloud is evolving naturally. There’s lots of healthy competition and a fundamental and underlying momentum behind the new cloud. Many “greenfield” applications (e.g., Animoto) are going to the cloud and probably wouldn’t exist in their current form without cloud services.
The virtualized data center, on the other hand, is heavily laden with future visions, and vendors are scrambling for position. The big food fight emerging is being catalyzed by the popularity of VMware, which is owned by EMC, the storage company that smartly purchased VMware for $635 million in 2003. Today, VMware is worth $15 billion. VMware is pretty much at war with everyone except EMC, Cisco, and a few integrators.
The virtual data center is a crapshoot and needs tested, validated configurations to work.
Enter today’s VCE announcement. This will make it easier for customers to adopt virtualization. It also will set off a chain reaction, as several VMware partners and allies have been left out of this announcement, including IBM Corp. (NYSE: IBM), Hewlett-Packard Co. (NYSE: HPQ), Microsoft Corp. (Nasdaq: MSFT), Dell Inc. (Nasdaq: DELL), NetApp Inc. (Nasdaq: NTAP), Hitachi Ltd. (NYSE: HIT; Paris: PHA), BMC Software Inc. (NYSE: BMC), CA Inc. (Nasdaq: CA), and virtually every other data center vendor except a number of IBM’s system integrator competitors (e.g., Accenture ).
The virtual data center is becoming a high-stakes battle for dominance of a large declining market where stacks are integrating; whereas the new cloud is a fight for dominance of a small but fast-growing market with vendors carving out a position at each layer of the value chain (e.g., storage, compute, platforms, applications, services, and clients).
While the new cloud and the virtualized data center share the same name (cloud computing) the differences are much greater than the similarities.
— David Vellante spent 15 years at IDC and is a founder of The Wikibon Project. He can be reached on Twitter at @dvellante.
The hybrid cloud (private cloud where you're moving workloads between internal and external clouds) is on the roadmap of VMware. IMO it will take a decade to get there. Getting data into the cloud is one thing. Getting it back reliably and quickly is another. But there are many more significant issues, including visibility on performance, automation of data and application movement, security, policies, governance, backup and the list goes on.
The reality is the data center world is one big heterogeneous hairball that unfortunately cloud marketing hype will not unravel.
Don't get me wrong. I love VMware's vision. It's fresh, bold, ambitious and interesting. It's great for the industry to have such a lightning rod. It's just the reality is it can only touch a small subset of applications today and it will take many years for VCE and others to prove that this model can be applied broadly in the data center.
And in the meantime, companies such as IBM, HP, Oracle and Microsoft will perpetuate heterogeneity because of their strong relationships and proven track records. Customers won't risk those relationships without very careful consideration.
There might be a third option in there -- the "Hybrid Public-Private" cloud -- that mainly serves one customer, but also has capabilities to attend to a handful of trusted partners/clients.
Or maybe the "cloud" will be even more amorphous than anyone suspects -- and everything will fall under a huge blanket cloud that encompasses any aspect of large scale computing. And the cloud will just be defined by how it's used...
Thanks for the excellent post Mike. You are right...the new cloud players don't covet a data center presence-- at least not today. There's too much growth in the stratosphere to come down to earth. The traditional data center crowd on the other hand don't want to be left off the cloud marketing train. In fact the day after the Berkeley paper came out essentially dismissing the VDC as not cloud, the spin masters in the traditional world hit the streets to stake their claim in the cloud and attach the cloud moniker to their solutions.
Your post was very insightful, and helped me gel what has been growing in my mind for a while now in the directions that the amorphous cloud seemed to be headed in.
I'm not sure that there are just two directions--time will tell that tale--but it is clear that the cloud announcements of the last year do seem to have two distinct rings to them--one is that any app or service from your data center could be moved into the cloud (a bold claim still) that has plodding progress by a few big players, and then a furious amount of activity by a bunch of new players that don't seem to be courting IT at all--in fact, they at times seem to be anti-corporate IT. This group has evolved into a set of ecosystems of their own.
The needs of the corporate data center, tasked with moving an app or service out of the data center into the cloud is different than that of the entrepeneur looking to build a service that would only be consumed from the cloud. The increased connectivity to a data center app afforded by the clould can both be a blessing, and a curse. The blessing could come in the form of increased access to the app by legitimate users remotely. These might be employees, or they might be "partners". The curse is that this is one more set of security nightmares for the IT wonks that can't even get us to change our passwords without draconian methods being employed.
If on the other hand, I'm building a service or application that take a set of loosely coupled servcies on a cloud-based bus--and the benefit of the application increases by having it be more openly available, then my needs are different.
I think the VCE Squad (insert your own lower case i) announcement is a key moment in the development of what David calls the private cloud. For in that world, clean interfaces, well understood and documented, make the transition easier for traditional "data center" applications--kind of like wrapping web services around a mainframe app to serve up a portion of browser integrated "page" sourcing data from multiple systems. Don't forget--Cisco UCS includes blade servers and host adapters for the fabric as well as the fabric itself. As David pointed out, IBM, Oracle, Hitachi, and the rest are all working their stories as well. The VCE announcement has some franchise names working in their favor; VMWare, EMC storage, and Cisco fabric and securty.
The Oracle postioning and product development/purchases about Grid and virtualization have been indicating their intent for quite awhile now. So I agree with you--they are to be watched--and for some, to be courted. Same with Dell, I would think.
Now that I see the new cloud for what it is, I can also see why the announcements are more frequent and more energetic. This is no longer an IT sell, this is appealing right to the parts of the business that are looking at creative ways of doing things that traditional IT by its very charter is ill-fitted to fulfill--linking us to Mary Shacklett's post on IT & Marketing cooperation.
I'm also curious to see what/if the new clould also begins to fragment as things go forward.
Hi Gaja...Thanks for weighing in...I think we disagree on all points except that Hyper-V has great potential.
*IMO the data center will take a decade to get to a point where you can move stuff from private to public cloud and back securely...i.e. the definition of private cloud as put forth by EMC/VMware and Cisco. Public cloud is there today-- amazon, azure, google, salesforce, etc.
*Virtually all of VCE's major competitors have responded and are in defensive mode. They are threatened...it is crystal clear. For example...Cisco is now a server vendor with UCS with Intel's full support. It's clear intent is to capture share in application and database markets.
*IMO Oracle's acquistion of Virtual Iron, its pending ownership of Sun and its extremely strong position in database/middleware and applications puts in in a much better position than anyone to control how/when/where database-based applications get containerized.
*My point about the 'true cloud' versus the data center wannabees is they are two completely different worlds. The 'true cloud' vendors aren't going after the data center but clearly VMware covets the market being pursued by Google, Amazon et al.
So we disagree, right? That's what makes good blogs!
I think Virtual data center is where private cloud is going to mature and public cloud is still trying to define what it is. VCE announcement may not be that bad for any other partners because VMWare, Cisco and EMC does not sell OS, Middleware, application servers or database servers. So still there is a lot of money to go around. It may rub Microsoft on their Hyper-V but MS is strongly positioned to grow in mid and small market's virtualiation. Oracle due to Sun's acquisition may never be able to get their virtualization act out.
On the flip side virtual data center maturity should make enterprises get comfortable on public cloud adoption. I dont personally see this as two different camps competing, but collaborating for greater market share.
Interesting comments Terry. I'd say: 1) I think competitors will have to break more than a sweat. VCE has set up a JV, they are spending millions on test and integration, they are spending on a services infrastructure that is a single interface to clients. This is in the boring but important department. But non-trivial.
IBM can do it. HP can do it. Others can to..but EMC and Cisco are stacking the VMware deck in thier favor. I'd be a bit upsset if I were a big VMware partner.
Polarizing? EMC/VMware/Cisco just declared open war with IBM, HP, Oracle, Microsoft and several others. It had a big party and invited 3 major players and some non-IBM integrators.
We've got Cisco, EMC and VMware adding a heavy dose of virtualization to the cloud equation, and then everybody else? Dell, Hewlett-Packard, IBM, Microsoft, NetApp and Hitachi won't exactly have to break a sweat to match the capabilities here, right? Is this new 3-way alliance really so polarizing?
The ThinkerNet does not reflect the views of TechWeb. The ThinkerNet is an informal means of communication to members and visitors of the Internet Evolution site. Individual authors are chosen by Internet Evolution to blog. Neither Internet Evolution nor TechWeb assume responsibility for comments, claims, or opinions made by authors and ThinkerNet bloggers. They are no substitute for your own research and should not be relied upon for trading or any other purpose.
Lately, I've been hanging around some cloud technologists, cloud service providers, virtualization customers, and security practitioners. I've been asking a lot of basic questions, trying to understand when and how cloud computing/virtualization will be ready to support any application or workload. UPDATED 2/2 4:00 PM
In a shocker of a news flash, the Chinese government has flatly denied any involvement in the cyber attacks on Google (Nasdaq: GOOG) and other Websites.
No matter what you call it -- virtualizaton, cloud computing, or "The Big Switch" -- a return to centralized computing is a widespread trend in our industry. And the general consensus is this movement will continue for a decade or more.
Last August at VMworld, VMware Inc. (NYSE: VMW) CEO Paul Maritz told me in the hallway, “We’re at war with everybody.” He was being flip, but this was no joke. The company had just gone out and paid $400 million-plus (about 20x revenues) for Springsource, an open-source Java framework developer.
Smarter Collaboration: How to Thrive in a Challenging Business Environment Market conditions are changing faster than ever, and organizations need to improve their agility and adaptability in order to provide better service and improve processes. The ability to work with customers, business partners, and employees as effectively as possible - while at the same time holding down costs - is a key to success. READ THIS eBOOK
your weekly update of news, analysis, and
opinion from Internet Evolution - FREE! REGISTER HERE
Wanted! Site Moderators Internet Evolution is looking for a handful of readers to help moderate the message boards on our site as well as engaging in high-IQ conversation with the industry mavens on our thinkerNet blogosphere. The job comes with various perks, bags of kudos, and GIANT bragging rights. Interested?
To save this item to your list of favorite Internet Evolution content so you can find it later in your Profile page, click the "Save It" button next to the item.
When Reiter gets incensed over incompetent Verizon FiOS order-taking and support, he broadcasts it via Twitter. Did it do any good? How should your company offer Twitter support? Watch this for all the answers.
Cloud computing is being dampened by the lack of local application support for offline use. Google's partnership with open-source should encourage it to build tight integration between Google Docs and OpenOffice, and thus boost the cloud and counter Microsoft at the same time.
Evidence shows that you can tweet too much. Sites and services like Twitter and Facebook are a good place to reach your audience, but think quality over quantity.
Telcos are falling over themselves to launch app stores – but are the app developers listening? Most telcos will need to do a lot more to engage their attention.
Research shows that the youth of today like Facebook – but not blogging or Twitter. Does that mean Facebook has won, or just that it's not yet out of favor? Will all the services we see today fade into Ovaltine-or-Wheaties status in just a few years?
What kinds of companies are doing the most innovation in the data center? Turns out it's midtier enterprises that are taking the "Just Right" approach.