Why Billion Dollar Red Hat and OpenStack Need to Dance

On Februar 29, 2012, Red Hat’s fiscal year came to a close and they are expected to cross an important milestone; becoming the first billion dollar commercial open source software company.  Whether or not you believe they are the first open source software company to cross this mythical threshold is inconsequential, the fact is Red Hat has done it.  With all my sincerest respect and admiration, I tip my “red hat” to this historical accomplishment.

With all due respect to other Linux distributions such as Canonical (Ubuntu) and SUSE, Red Hat is the de facto standard for Enterprise Linux.  They have a reputation for building a quality product, have a stable of certified applications from leading ISVs, maintain a “Cisco-like” army of certified professionals, and provide long term support for their products.  Unlike the early years of their business, Red Hat’s biggest threat does not come from a new operating system challenger ala Microsoft; it comes from virtualization vendors with all eyes on VMware, Microsoft, Citrix, and Oracle.

The good news is that Red Hat foresaw this threat and purchased Qumranet in 2008, which created kernel-based virtual machine or KVM.  The bad news is while VMware grew to a virtualization powerhouse, it took Red Hat until January 2012 to release a real challenger within Red Hat Enterprise Virtualization 3.0 (RHEV3). With this release, the next chapter in Red Hat’s history is unfolding within the Cloud era.

Meanwhile, OpenStack is nothing short of an amazing story of the power of open source and of community.  In July 2010, Rackspace and NASA jointly launched OpenStack and less than 2 years later OpenStack has the backing of over 150 companies with names like Dell, AT&T, HP, Citrix, and more.  Additionally, you’ll find the likes of Canonical and SUSE, but Red Hat is noticeably absent.  However, is Red Hat really that far away?

All one has to do is open OpenStack’s Administrator Manual and you will find instructions on “Installing OpenStack Compute on Red Hat Enterprise Linux 6”.  Also, looking closely at the list of contributing companies you will find Gluster (one of my favorites), a project acquired by Red Hat on the list.

In the battle for virtualization supremacy, OpenStack is a vital weapon against the competition.  Sure, Red Hat has Aeolus and Deltacloud, but what would the world look like with a RHEO (Red Hat Enterprise OpenStack) edition?   Wouldn’t such a release accelerate OpenStack’s rise in the Enterprise while opening up a new revenue source for Red Hat?

Before any of this can happen, Red Hat and OpenStack must dance.  Sure, there are reports that early on Red Hat was invited by Rackspace to join OpenStack but they refused due to its governance model.  However, things have changed as Rackspace has transitioned management of OpenStack to an independent OpenStack Foundation with a defined mission and structure.  Can Red Hat and OpenStack unite under this new model?

Perhaps the final piece of this puzzle will end with a Red Hat acquisition within the OpenStack ecosystem.  Without naming names, there are at least 2 attractive take-over targets that would give Red Hat the development expertise and OpenStack credibility to be a force.

As John Lennon famously wrote and performed in “Imagine”; “You may say I’m a dreamer, but I’m not the only one, I hope some day you’ll join us, And the world will be as one.”

Red Hat and OpenStack, let’s dance!

Originally posted on http://blog.zenoss.com

Advertisements

Floyd’s Law: Open Source vs. Proprietary Software

As the pace of innovation continues to accelerate, it is increasingly impossible for legacy software vendors to maintain pace.  Professional services organizations are pushed to the brink as they attempt to fill product gaps only to find that they are further and further behind the innovation curve.  Customer frustration is increasing as these projects never end, product innovation never comes, and maintenance costs continue to increase.

Open Source, free of the legacy baggage and bureaucracy of their traditional competitors, is the only model that can keep pace with the accelerating rate of change in the industry.  In fact, Open Source is the disruptive force that continues to break-down legacy paradigms and offer new and disruptive solutions.  As commercialization of Open Source is inevitable, the key is remaining true to the principals of open source while providing customers the innovation and value they desperately desire.


Cloud Wars: Cisco Invades OpenStack

Simply put, Cisco is an amazing company.  Love them, hate them, fear them, or fight them, but always respect them.  While other large companies such as IBM and GE have “reinvented” themselves, few have done so prior to having a profound downturn in either market share and/or stock prices.

With an innocent blog post by Lew Tucker, Cisco VP and CTO of Cloud Computing, Cisco has invaded OpenStack under the guise of Networking.  Remember, OpenStack was founded by NASA and Rackspace and currently has over 45 members with the mission of providing open source software to build public and private clouds.  However, none of OpenStack’s members have the shear size or market power of Cisco.

In his post (http://blogs.cisco.com/news/cisco-joins-openstack-community/) Lew writes:

In our view, dynamic provisioning of the network and network-based services is an essential element of cloud computing…

…To achieve this, we believe that it is best to join with others from across the industry to work on open technologies and that open source is the ideal way to reach developers and learn from the community…

My Take, Cisco is spending billions of dollars to insure their continuing dominance in networking and Cloud computing.  By joining OpenStack, Cisco gains visibility into OpenStack’s interworking as well as the ability to influence the direction and speed of the project itself.  As an added benefit, Cisco will learn from the community while having the ability to reach a set of talented developers that otherwise may never have engaged with Cisco.

Clearly, Cisco understands how to build complex partnerships across competitive lines.  While VMware has vCloud, do they not work with other server vendors?  Would EMC not sell a SAN to an HP customer?  As a server vendor, Cisco is learning that choosing neutrality over products has its benefits especially when it comes to software.

While OpenStack is “hot” and an interesting project, they have their competitors as well with more coming. It remains to be seen if,  “a collection of open source technologies delivering a massively scalable cloud computing operating system” is supportable and useable by mainstream Administrators and Enterprises.  Perhaps this is what VMware is betting on with their vCloud solution.

One final note, where are the Operating System Vendors in this fight? Yes, Ubuntu is currently packaging applications like OpenStack and Eucalyptus but we need an integrated Cloud Operating System, not simply a collection of applications.  Microsoft, Red Hat, Apple, anyone….

Cloud Wars: Dell Fights Back With UEC

While Verizon is acquiring Terremark and Time Warner Cable, yes TW Cable, is acquiring NaviSite, Amazon continues to disrupt the industry with their 12 plus Cloud offerings.  The more EC2 grows, generated $220 million in 2009 with predicted revenue of $500 Million in 2010 and $750 Million in 2011, the more it validates that customers are willing to transform their purchasing behavior from hardware devices to compute nodes.

Meanwhile, Enterprises are struggling with virtualization and virtualization stall with the impending reality that they must operate within a Cloud model.  Here lies VMware, the dominant x86 virtualization provider, as they have a complete set of products and 3rd party certified partners to help their customers go virtual.  Let’s face it; ESX/ESXi and vCenter are excellent products.  Additionally, VMware has introduced vCloud and vCloud Express “VMware Power. By the hour.” Essentially, this technology allows Enterprises to build a private cloud and Service Providers to build public clouds and to provide hybrid cloud offerings.

Of course, this pits Amazon’s Cloud Offerings, which are not built with VMware’s technology, against VMware and some of their most powerful partners.  Amazon utilizes the Xen hypervisor along with other customized/internal solutions.  Understanding that VMware is the dominant Enterprise x86 virtualization technology, Amazon has introduced VM Import.  VM Import allows Enterprises to easily migrate VMware Guests (VMDK) into the Amazon EC2 Cloud.

However, what if I want to create a private EC2 within my Enterprise?  Along comes Dell’s Ubuntu Enterprise Cloud (UEC) infrastructure solution. Dell UEC combines the power of Dell’s server hardware with the software of Ubuntu Linux and Eucalyptus providing Enterprises with the same virtual machine images and management APIs that Amazon uses for EC2.  Well it is not exactly EC2, as some will argue that Eucalyptus is not a full implementation of the EC2 API, and it is a matter of fact that Amazon has plenty of additional customized internal tools/systems that make EC2 a reality.  Not to mention that EC2 relies on the Xen hypervisor while UEC utilizes KVM virtualization.  All in all, it’s a great start.

As always, Dell has published an excellent UEC Reference Architecture White Paper for UEC Standard Edition.  This begs the question whether or not Dell will offer Enterprise and/or Service Provider Editions of UEC.  In any case, Dell now has a visionary offering that they will be able to evangelize to their current customers and prospects.  In fact, as UEC matures, Dell is sure to add elements of their entire product portfolio; namely Compellent storage equipment, more powerful server platforms, and perhaps networking/storage hardware via their partnerships with Juniper, Brocade, and others.

One last thought, Dell has incredible flexibility in creating unique cloud offerings via simply changing software and hardware partners.  For example, offering a solution based on Red Hat with Delta Cloud or perhaps a secondary UEC offering that utilizes OpenStack.  This flexibility also translates to Dell’s Open Management philosophy, which is sure to attract additional software partners thereby creating a UEC partner ecosystem.

Can Cisco Eat their EMC and Have Their NetApp To?

With 2010 nearing a close, could Cisco be contemplating another major acquisition to complete their next generation datacenter portfolio?  The last glaring hole within Cisco’s portfolio is their reliance on outside vendors for storage solutions.

Over the past few months, Cisco has patiently watched as HP purchased 3Par, EMC purchased Isilon, and Dell is acquiring Compellent.  Meanwhile, EMC’s arch nemesis NetApp continues to grow and innovate in a tough economy.

Further complicating matters, is Cisco’s reliance on the VCE, a partnership between VMware, Cisco, EMC, and Intel.  It is no coincidence that the current Vblock VCE Reference Architectures specifies EMC storage offerings (CLARiiON, Symmetrix, and Celerra).

Not to be left out of the party, NetApp entered into  ‘collaboration’ with Cisco and VMware creating FlexPod that delivers ‘leading computing, networking, storage, and infrastructure software components’.  It seems that Cisco isn’t the only one hedging their bets as VMware exerts a rebellious streak against their parent (EMC).

Cisco’s future hinges around UCS being adopted as a true next generation computing platform without legacy baggage.  Cisco did not go to war with HP while potentially jeopardizing their relationship with IBM only to be saddled with the competing interests of three large companies.

In the past, I have speculated that Cisco should simply purchase EMC thereby owing a majority stake in VMware.  However is NetApp a better choice?  After all, does VMware need to maintain a ‘Microsoft’ level of independence from the server vendors?  Would HP, IBM, Dell, etc. be inclined to sell a product that lines the pocket of Cisco?

Only Chambers (ok perhaps Ellison as well) would be as bold to acquire an enemy of one of their strategic partners.  By acquiring NetApp, Cisco would be able to offer innovative solutions such as storage blades for UCS or even accelerate the adoption of FCoE.  Imagine a new Cisco Architecture with Cisco UCS, Cisco Nexus, Cisco MDS, Cisco FlexPod, and Cisco Management with the availability of VMware, Citrix, Red Hat, or Microsoft virtualization.

In the end, Cisco could offer a true end-to-end solution as they continue to lead within the edge and core routing markets with near dominance in the switching market.  Furthermore, Cisco would stand alone as the only integrated next generation data center provider that does not develop or sell enterprise class applications such as SAP, Oracle, Microsoft, etc.  In effect, they become the Switzerland of computing against their rivals.

The only question is how long will Cisco be able to ‘Eat their EMC and have their NetApp to’? Don’t look now, but perhaps Larry (Oracle) will crash this party and make the decision for then.

Oracle Takes Dead Aim At VMware’s Vision

Still wondering why Oracle purchased Sun?  Day One of Oracle’s OpenWorld 2010 cleared up at least one reason; Oracle has its own “stack” and it does not include VMware. 

When Sun was originally purchased by Oracle, my attention immediately fell to Sun’s virtualization assets and engineering talents.  Before the acquisition, Sun was amassing an arsenal of virtualization and management assets including xVM, VirtualBox, and Solaris.  If you factor in hardware development and JAVA, then Sun had everything they needed to “change the world.”  That is, everything except a track record for translating engineering into revenue.

Love him or hate him, Larry Ellison has no such issues.  His track record speaks for itself as Oracle has an uncanny ability to execute.  With the launch of Exalogic Elastic Compute Cloud, Oracle has signaled to the market that they are ready for a fight.  While most believe Oracle is gunning for Amazon Web Services, I believe hidden in their messages and jabs at IBM is their true target of public, private, and hybrid clouds ala VMware.

VMware unleashed their vision of the future at VMworld 2010 that included vSphere, vDirector, vCloud, vFabric, SpringSource, and more.  What’s missing?   Oracle would point to VMware’s ratio of vision to products, their lack of owning an operating system, and their dependency on third parties to deliver server power (I’ll give them storage as EMC is VMware’s parent company).  Oracle’s vision is unique in that they control the entire Cloud stack using proven technologies and deployments; unleashing the potential of Sun hardware, JAVA, and Fusion.

Exalogic Elastic Compute Cloud has a few things going for it:

  • Power – Scale Out and Scale In  
    Cores 96 to 2880, SSD 256GB to 7.7TB, RAM 768GB to 22.4TB, and SAS disk 40TB to 320TB
  •  Applications – JAVA and Fusion
     Oracle’s Applications as well as others that run on Oracle Solaris and Oracle Linux

Although, I’m not thrilled with Oracle’s reliance on InfiniBand, it makes sense given Sun’s product portfolio and expertise.  Also, we need to learn more about how you manage this system including orchestration via business process management solutions.  However, this is a great start for Oracle. 

One last thought, Oracle took a subtle jab at VMware, EMC, and Cisco when they proclaimed, “Run 1000s of existing applications” and “No Certification Required.”  Perhaps Ellison should not be picking a fight with Mr. Chambers at Cisco.  For the common denominator of vBlocks (VMware, xBlocks (Citrix), and rBlocks (Red Hat) is UCS and its momentum may be unstoppable.

Three Cheers: Cisco Unleashes a UCS Surprise with ‘xBlocks’

With today’s webcast and subsequent announcements, Cisco showed their continued focus and commitment to the UCS platform.  The results are breathtaking; a new B230 M1 ½ blade with 16 cores and 2048/4096GB of memory, a new Nexus 5500 series that doubles the port density of the previous generation 1U models for up to 960Gbps of throughput, and virtual appliances for both Virtual Security and WAAS.  Perhaps the most interesting announcement of the webcast focused on VDI (Virtual Desktop Infrastructure) sporting a partnership between Cisco, Citrix, and NetApp that I’m dubbing xBlocks.

Like vBlocks, xBlocks maintain their own Reference Architecture that was jointly developed by Citrix and NetApp and validated by Cisco as a CVD (Cisco Validated Design).  Unlike vBlocks, xBlocks shed VMware for Citrix and offer an implementation based on Citrix’s XenDesktop infrastructure.  Additionally EMC is swapped for NetApp as they are the perfect non-competitive complement to Citrix.

While Cisco is heavily invested in VMware, this announcement demonstrates Cisco’s desire to broaden the UCS audience.  Of course, this isn’t the first time Cisco has ventured from VMware’s path as in June of 2010 Cisco announced a strengthening of their relationship with Red Hat and their KVM hypervisor.

Finally, Cisco confirmed 1700 UCS customers worldwide with 200 Unified Computing Authorized Technology Providers (ATP).  This means that Cisco almost doubled the total number of UCS customers from the quarter before.  If anyone doubted Cisco’s ability to disrupt the server market, then these numbers clearly demonstrate that Cisco is succeeding at an alarming rate to their competitors.

Is Cisco blowing up the old notion that innovation only comes from start-ups or small companies?  Are Chambers’ course corrections leading to a right hand turn and the complete transformation of an industry giant?  Is HP focused on this market or are they too busy buying and integrating companies?  Is Larry pacing his mansion contemplating how to leverage Sun to get into this fight?

Three cheers to Cisco:  innovators without baggage, partnerships without exclusivity, and a platform that is breathtaking.  What’s next, I can’t wait!

%d bloggers like this: