Thin Margins, Change and Differentiation

 

As we watch the explosion of cloud events, press releases and panels is anyone getting a sense of déjà vu all over again?  It wasn’t that long ago when Webvan was going to transform the grocery business with new technology and processes. 

 

After a flurry of announcements and expansions we learned that innovation could be tough in low margin/high volume service businesses.

 

Re-architecting highly segmented enterprise data centers (because of security, regulations, importance of data and access levels, etc) into dynamic pools of on demand processing power isn’t an easy task.  Segments exist for important reasons.

 

For light SMB applications and x-as-a-service there is already a market.  The enterprise IT as a service questions are really about scale, margins, flexibility and security; that is, enough for small and medium-sized businesses.  Then there is a potential for more innovation as virtualization spreads and creates new markets for management and infrastructure software.

 

No doubt we’ll see service specializations with emphasis on app and vertical expertise and responsiveness.  Yet these new specializations could still be stuck between Amazon, Google and a myriad of established players with paying customers, service expertise and existing profitability… and the evolution of enterprise IT to the intercloud.

 

The intercloud revolution will take some time, especially when you consider intercloud challenges.  As soon as enterprises can convert scattered data centers into elastic meshes of processing power the business case for densely architected cloud startups will be disrupted.  The intercloud can chase low cost rates from one optimum location to another.

 

A host of infrastructure service providers are likely to get to cloud first and evolve into dynamic meshes before the startups (growing and having to innovate on thin margins).  While the advantage of concentration (the massive data center with dense blade servers) shifts with intercloud (the massive mesh of distributed data centers), the timing of when that shift happens will make all the difference for the thin margin, service players.

 

Today’s established service providers and enterprise IT teams are much more likely to innovate profitably, unless they make the wrong investments at the wrong time. Cloud technology is likely to experience multiple waves of disruption in early stages, and investing too early or too late in various approaches could have material consequences.

 

I moderated a Cisco panel last week at Cisco Live! And it was readily apparent that enterprise cloud still required plenty of work from vendors and enterprise IT.  You can expect substantial innovation in coming months, according to at least one panel participant.  For example, no one was ready to endorse a centralized versus decentralized architecture (a move to the powerful intercloud); perhaps it’s because the network isn’t ready for infrastructure 2.0 demands.  When it is it will make all the difference.

 

Another cloud question relates to goliaths Google and Amazon.  Amazon is probably in better shape from a classic business model basis.  It is leasing off-peak, existing capabilities with line extensions while Google aims to re-architect IT, perhaps from the ground up.  The Google strategy reminds me of Webvan to some extent, except with more cash, customers and options.

 

Remember the Webvan

 

Yet re-architecting enterprise IT (high volume, low margins, service excellence, etc.) still doesn’t sound that different from the recent attempts to re-architect the grocery business. The Internet home delivery sector that (I confess) pinched my dotcom portfolio comes to mind as we watch the flock of cloud players forming to reinvent enterprise IT:

 

“Among the Internet’s last players in the home-delivery sector, Webvan has seen the dot-com shakeout trim the ranks. Companies like ShopLink.com and Streamline.com have folded, and online convenience store Kozmo pulled the plug on its operations Wednesday.”

 

Greg Sandoval, CNET News, April 17, 2001

 

Google is in much better shape than the grocery upstarts ever were, yet they have an ambitious vision.  The most compelling cloud service provider story I’ve read thus far is a recent blog by Benchmark’s Bill Gurley.  He argues that Amazon’s model already leverages low margins and excellent customer support, two critical factors for cloud service success. 

 

Like the grocery story stores taken on by Webvan, Amazon is already entrenched in the category with core strengths that may otherwise be hard to assemble within the confines of a high volume (scale) and thin margin business experiencing technology transformations on a frequently massive scale.  Google certainly is better capitalized than Webvan and has a strong core business.  Yet is it strong enough?

 

Gartner’s Whit Andrews, quoted in another piece of CNET coverage offers sage advice to those who wanted to enter the online grocery business:

 

“Here’s a radical thought: The future of the online grocer market belongs to the grocery stores,” Andrews said. “They know the business, they can mix (sales) channels, and they can take their time.”

 

Melanie Farmer and Greg Sandoval, CNET News, July 9, 2001

 

Amazon will likely beat the CloudVan model. After Amazon, Google has perhaps the best chance to win, versus a host of cloud start-ups betting on scale, service and thin margins, etc.  Unlike Webvan, Google has a highly lucrative and complementary business.  They have been accumulating top talent and generating incredible buzz.

 

The rest of the cloud pack survivors will likely come from the pack of successful legacy service providers who make evolutionary investments and serve specialized applications or market niches.  Again, they’re the incumbents used to the rigors of tight margins and exemplary service.

 

More Surprises from Cisco?

 

Recently Cisco started talking up the idea of its own cloud app offerings.  Suddenly the world of servers, computers and applications starts looking more like a rugby scrum than well-segmented complimentary markets. 

 

This leaves Google cloud and legendary rival Microsoft in an awkward position.  Microsoft is caught between Google and the rise of netbooks and increasingly powerful alliances between VMware and Cisco and IBM and Juniper, etc.  They have a massive footprint that yet appears to be under siege from almost every angle.

 

Tech-savvy Google, on the other hand, could be triggering a transformation that benefits nimble, visionary incumbents versus the fresh crop of cloud service provider startups.  The world of IT is breaking fresh ground these days, mostly due to the three horsemen: virtualization, cloud and netbooks.  Fortune will likely go to the bold, innovative and nimble.

The intercloud is the most interesting development in cloud computing (and perhaps the data center) since the migration of hypervisors into production environments.  It’s a model for driving incredible scale and efficiency from IT systems and hardware, and it can be architected for both public and private clouds.

 

The intercloud turns a global network of clouds (or virtualized data centers) into a cohesive, elastic and flexible mesh of on-demand processing power.  It shifts the balance of IT power from centralized to distributed architectures by allowing enterprises a wider range of on demand IT service delivery choices. 

 

For example, services could be delivered from one location over another because of short term differentials in power and/or labor costs.  It would also give enterprises more viable options for dealing with localized tax or regulatory changes.

 

The intercloud doesn’t yet exist, however—It has at least one missing piece: the automation of manual tasks at the core of the network.  The intercloud requires automation of network services, the arcane collection of manual processes required today to keep networks and applications available.

 

Until there is network service automation, all intercloud bets are off.  This gap between the delivery of cloud services today (via higher VM density and centralization) and the arrival of the distributed intercloud is the next period of stability before the next IT disruption.   Because of the low margin/high service (and high CapEx investments) involved in building scalable, elastic cloud services, the size of this intermediate cloud period will make all the difference for service providers and startups.

 

The Intercloud’s Net Effect: Read the rest.

I’m moderating a panel July 1 on the data center of the future made up of a great team of executives from Cisco, Oracle, NetApp, VMware, EMC, APC, Emerson and Panduit.  Hoping to catch up with a few friends after the panel.  Drop me a line if you’re attending.

I don’t know who coined the “intercloud” term, but it is a big step forward in defining the enterprise cloud endgame.  See this Cisco blog and recent intercloud preso.

 

As sexy and compelling an idea as it is, the Intercloud would probably today be a bane according to Mark Masterson, who created a brilliant presentation on enterprise cloud security.  In essence he talks about the unintended consequences of larger computing systems with more complexity.

 

I think the gap between the intercloud promise and reality is the network.  That’s why it makes great sense for companies like Cisco to begin educating the market about where it needs to go, and ultimately the automation and management required.  The vision will help to champion the technology required.

 

That means it is time to talk about the challenges of intercloud, before we get too carried away with the promise.  Everyone probably agrees that it will be the most powerful, cost effective and compelling form of cloud computing for the enterprise.  Yet according to Cerf, today’s Internet isn’t ready for the simpler stuff.

 

“One of the most critical needs is authentication, Cerf said, and he told the crowd at a TechAmerica gathering Wednesday that anyone who performs transactions over the Internet — which is everyone — “should be deeply concerned about that technology.”

The lack of authentication is pervasive and is even a problem in simple cases, such as authenticating entries in the domain name system, he said.”

Patrick Thibodeau, Computerworld June 11, 2009

 

There are still issues with management and virtualization, as Denise Dubie suggested in Network World recently:

 

“A majority of IT departments are deploying virtualization, but still most don’t feel comfortable with the tools and technologies they have in place to manage application performance or troubleshoot problems in the virtual environment, according to recent survey results.”

- Denise Dubie, Network World, May 2009

 

More specifically, the gap between today’s network and the intercloud will need to be addressed by more automation and integration around some of the more mundane stuff at the core of the network, like DNS and IP address management, which perform a key role in making sure that network traffic reaches the right place.

 

You can catch some of the buzz about infrastructure 2.0 from the recent Future in Review Conference.

You can get the slides for today’s DNSSEC webinar with Dan, Cricket and Scott here.  The version with audio and slides will be available at the dns security center and the DNSSEC solution section next week.  You can follow Cricket’s answers to questions asked about DNSSEC at the new Cricket Liu blog.

A few months ago I blogged about the Three Horsemen of the Coming Network Revolution.  One of those “horsemen” was the netbook computer.  Now Microsoft wants to rename it. Maybe it has something to do with the impact of the netbook on the hard drive PC market and the powerful relationship with robust, pre-loaded operating systems and software.

 

The netbook could decouple that relationship, hence the silly idea of a name change and a recent JavaOne Conference headline: “Ellison: Java-based Netbooks on Sun’s Horizon.”  Netbooks could disrupt the playing field by transforming how endpoints interact with the cloud versus the hard drive.

 

That won’t be an overnight change, but it is one of notable proportions.  Various service providers (and endpoint players like Apple) can offer suites of applications architected for cloud demands; flash memory players can compete with the hard drive posse; and new cloud architected semiconductor chips can unhinge the power of legacy expertise and architecture.

 

We’re possibly entering a new and exciting era for network infrastructure that will force innovation out of what has become a tired network and  a stagnant pre-load PC industry today docked in the port of commoditization as the cloud storm approaches.  New competitors, new potentials and new relationships will have more opportunity than ever; and the old guard needs to rethink the new infrastructure 2.0.

 

This is a bigger challenge than a netbook renaming project.  It will require vision, agility and momentum.  For some players these are not core strengths.

After Interop I moderated a Future in Review (FIRE) panel and had the pleasure of asking executives from Infoblox, Cisco, F5 Networks and VMware about their perspective on infrastructure 2.0.  The first part of the panel talked about the problems with today’s static networks, including rampant manual labor and outdated management tools; the second part explored the economic consequences of static infrastructure for business and jobs.

 

FIRE Infrastructure 2.0 Panel – Photos copyright © 2009 by Sandy Huffaker Jr

 

Infoblox’s Ness (blog author) and Kagan, with VMware’s Thiele, F5’s Giesa and Cisco’s Gourlay at the Future in Review Conference May 21, 2009 

 

For application and infrastructure as a service (now forms of cloud) the cloud computing vine has indeed been fruitful.  Yet with much broader and more complex enterprise IT services there are substantial barriers, which we’ve discussed before in 3 Major Barriers.  How those barriers are tackled will mean all the difference to a community of public and private companies.

 

The Fire panel marked the first time that three leading networking and virtualization companies have shared the stage with the same lament: it’s time to automate network infrastructure.

 

Cisco’s Gourlay predicted that cloud deployed on ill-prepared networks could break the Internet.  F5’s Giesa talked about the need to allow application delivery policies to follow mobile systems to ensure proper delivery.  VMware’s Thiele discussed new demands on data center managers which will force network shortcomings into the open.  Infoblox’s Kagan talked about the core need for core network service automation as a starting point for Infrastructure 2.0.  You can view the entire panel discussion here via Vimeo.

 

During the introduction I commented that: “networks today are run like yesterday’s businesses and factories, with legions of …clerks…”  I drew a parallel between the rise of automation and supply chain in manufacturing and the coming impact of cloud on IT service delivery.  Cisco’s Gourlay coined the term “just-in-time IT services” to describe where cloud is taking IT and the parallel ramifications for companies, jobs and the wealth of nations.

 

The core cloud/network problem: There are layers of outdated practices that today are extracting volumes of micro-tolls from enterprise networks, simply because they’re usually informal, ad hoc and scattered among units.  These practices need to be automated and integrated with each other or the fruits of virtualization and cloud will be minimized in darker clouds of expense, delays and outages.  Networks need to be intelligent and able to keep up with the changes ushered in by the automation of systems enterprises are carrying out today. Simon Bisson summarized it well on this IT PRO blog:

Perhaps the most telling piece of the puzzle was one simple phrase: “We need to stop treating IT pros like Victorian file clerks”. It’s a statement that hit home – we do treat our IT pros as glorified clerks, waiting for them to do things by rote. What we really need is an automated infrastructure that flexibly configures itself to deal with the tools, applications and workloads we need to use every day.

Pull apart all the different definitions from all the vendors out there and that’s what Infrastructure 2.0 boils down to. It’s a world we really need to build – if only to show the world just what value IT really brings to business.

- Simon Bisson, IT PRO, May 27, 2009

 

Today I read about the Emerging Technologies Forum hosted by Integrated Archive Systems on June 24 in Mountain View, California.  The list of sponsors includes Cisco, F5, VMware and Infoblox, along with NetApp, Microsoft and Sungard.   From the looks of it, I think that the I2.0 herd is continuing to form as other voices join the conversation.

 

Public Companies in the Infrastructure 2.0 Mix

 

If someone was interested in following public companies well positioned for the evolution to infrastructure 2.0 one would naturally include Cisco, VMware, F5 Networks, Citrix, IBM and Juniper.  Notably, Cisco and VMware recently won Best of Interop awards for their infrastructure 2.0 innovations.  Amazon, Google, and numerous service providers (including Sungard) could transform the IT services (especially for small and medium businesses) if they can drive the evolution of cloud services and applications and minimize unplanned downtime.

 

A Not-So-Wild Idea: OEM Netbooks from a Network Player

 

As the lines blue between traditional networking players and servers and services I think it is very possible to see a networking player (or even a phone handset player) enter the netbook space.  Imagine a netbook with logos from network vendors rather than processor vendors; after all, isn’t that what cloud is all about?  It’s the network and service “inside” that really impacts performance, availability, security, etc.

 

Netbooks could generate brand visibility and help drive more bandwidth demands for network vendors.

 

A Private Player of Interest

 

At Future in Review I also saw a cloud application from SIMtone that garnered its fair share of enthusiasm. SIMtone demonstrated a kind of cloud-based portable desktop that can be easily transferred (on demand) from one device to another.  I’m tempted to call it a virtual desktop service, for lack of a better term.  According to CEO Mario Dal Canto a virtual desktop can be accessed via hyperlink or even a widget.

 

Cloud computing may still be years away from strategic implementation in large enterprise data centers, but it is already casting a long shadow.  Its impact across IT is about to elevate a new crop of startups and perhaps invigorate a few familiar names in the tech industry.  And it is likely to rain on the fortunes of still others.

DNS Tools

If you are looking for DNS tools check out DNS test (or DNS freeware) at the Infoblox web site.  You’ll find links t plenty of tools, freeware and additional resources to help you manage DNS.

 

Interop

If you’re planning to attend Interop feel free to visit the Infoblox booth 1867, in the TNC demo (booth 869) or at the Cisco (1719) or Riverbed (1359) booths.

 

The Infoblox booth will feature demos of the new Infoblox IP address management (IPAM) solution as well as the new PortIQ™ appliance network monitoring solution.  The TNC demo will show how IF-MAP enables security for 5 distinct enterprise environments.  If you cannot see the demos live check out the new 2009 TNC Interop brochure.  Here is a sneak peek.

 

FIRE Conference Panel

I’ll be speaking on a panel about what can go wrong with cloud computing (at 1:40 in the Enterprise Cloud Summit).  Feel free to stop and say “hello” if you’re in the room.  On Thursday I’ll be on a Fire Conference panel with Cisco’s Gourlay, F5’s Giesa, VMware’s Thiele and Infoblox’s Kagan.  We’ll be talking about the evolution of Infrastructure 2.0.

 

DNSSEC and DNS Security

Infoblox is hosting a live webinar on DNSSEC on June 10 with Scott Rose, Cricket Liu and Dan Kaminsky.

Cloud computing and virtualization are promising more dynamic systems with unprecedented cost savings; and network industry leaders are promising more dynamic networks capable of keeping up with the increased rate of change (with these systems).  In order for the benefits to be delivered as promised IT will need to evolve from silos into multifunctional teams, and vendors will need to concomitantly embrace their partners like never before.  This process is already underway.

 

I recently addressed the three biggest barriers to cloud: security, network capacity and network management. These barriers will likely be addressed via new and cloud-strategic partnerships of various kinds as the silos of legacy IT converge into pre-configured containers blending multiple vendor offerings that can be scaled up and out to maximize flexibility and cost savings. 

 

Assessing the Leaders

 

VMware has been the most active of the virtualization players in addressing the new demands of virtualization security.  Whether Microsoft and Citrix continue to rely upon partners (or internal development) or make strategic acquisitions of one or more of the virtualization security startups remains to be seen.  Ultimately it will be their ability to address customer security demands that will establish their solutions as a clear path to secure, scalable cloud deployments.

 

On the network capacity side you have to be impressed with the way that Cisco’s UCS with its new cloud angle is evolving.  Network World certainly liked their Catalyst 6500 in a recent review.  At this point they seem to be ahead of the network switch pack; although rival Juniper has announced a partnership with IBM and more scalability and density for service providers.  Juniper’s historic core strength with service providers cannot be underestimated as packs of cloud service providers emerge.

 

F5 Networks earlier announced their enhanced software for managing dynamic infrastructure (or infrastructure 2.0- a term for a network capable of supporting virtualization and cloud).  Their expertise with application delivery gives them a potent point of leverage.  Cisco, Juniper and F5 all understand the implications of the coming sea change in IT and are shaping their offerings to build the roadmap for their customers.

 

DNS, DHCP and IPAM- The Front Line

 

In between today’s growing, increasingly complex and dynamic networks and the promise of cloud computing are layers of challenges driven at least in part by legacy IT.  IT services at the enterprise level have been managed by silos of networking teams, security teams, server teams, operations teams, data center teams, etc.  As VMotion is fully enabled, many of those silos will become costly and irrelevant. 

 

While applications and networks and systems are containerized into holistic and replicable offerings, the services that connect and coordinate and deliver them will require unprecedented automation as the silos dissolve and teams begin sharing resources and responsibilities.  As enterprises plan this evolution, IPAM (or sometimes IP management) will cross the chasm in the company of integrated DNS, DHCP, network monitoring and other core network services. 

 

Very few CIOs today are familiar with DNS or DHCP; expect that to change as IT begins planning its evolution to infrastructure 2.0.  They will become as critical to CIOs as maps are to generals.

 

Across IT these dedicated, integrated and automated DNS services will be the front lines as systems decouple from hardware and the importance of knowing the location and history of IT assets grows exponentially with increasing rates of change and complexity.  Today the spreadsheets and manual configuration requirements are already wearing network teams thin.  Mix in accelerating change and complexity and you have obvious operational breaking points.

 

As networks grow and become more complex their management costs escalate even faster.  Core network services, therefore, become critical to availability and security, in the same way that the LAN and WAN became critical components of new IT delivery strategies as enterprises moved employees to personal computers and networked services and then spread computing to branches, teleworkers and partners. 

 

Today the IT strategy list is much larger (including VoIP, wireless/RFID, NAC, Web and eCommerce/supply chain); the services are more critical.

 

Hence it is likely that CIOs and IT VPs will become increasingly aware of the connectivity intelligence challenges inherent with increasingly large and dynamic infrastructure and these new acronyms.  If they don’t, many of the consolidation benefits from cloud will be offset by the rising manual network management requirements tied to complexity and velocities of change, not to mention reduced network availability.

 

If you don’t see this coming check out the netbook revolution that is already unfolding, introducing the prospect of even more endpoint growth and ever more reliance on the network.  The enterprise cloud is under construction; and before cloud accelerating endpoint growth appears to be a predictable eventuality for most companies.

 

The success of the larger IT vendors who want to migrate their customers away from silos and into unified fabrics will depend on their ability to automate these manual tasks.  You can read more about this at “Clouds, Networks and Recessions”. 

 

As vendors aim for the clouds they’ll need to address more than security and throughput/capacity.  They’ll need to embrace automation with solutions which are integrated with DNS, DHCP and IP address management or IPAM.  CIOs overseeing these growing and increasingly strategic networks may soon become familiar with acronyms once considered too mundane for executive interest.  Those who don’t may end up wondering why things don’t go as smoothly as promised.

 

 

I am a senior director at Infoblox the leader in DNS appliances and IP address management You can follow my comments in real time at www.twitter.com/archimedius or catch the conversation as it happens at www.infra20.com.

As I mentioned at 3 Major Barriers to Cloud, the transition from tech services silos to unified cloud computing fabrics will trigger rapid growth and integration in network monitoring and dns appliance solutions.  Integrated, automated, IP address management will be consolidated into Grids with DNS and DHCP and network monitoring solutions to deliver integrated views and management of critical network resources.

 

Cloud will place more demands on the network and drive enhanced business cases for network automation and network management.

 

Infoblox just announced its PortIQ appliance, which can be integrated with its IPAM software.  You can check it out at Interop next week, booth 1867.