NetApp’s Cloud Journey - Continued

In the beginning of our Journey, I covered Cloud origins and history, as well as NetApp’s definition of our Cloud Market opportunity.  Key lessons around not competing against your best customers and the requirement for a Cloud Ecosystem and implementation scope were also reviewed.

This blog will continue the discussion, building on the concepts from my prior blog in order to complete the introduction of NetApp’s Cloud Strategy.

#6 – Understand your true competition

The cultural (i.e. Millennials) and economic pressures at the end of this decade are creating an undeniable tension within Enterprise IT departments.  Public Cloud Services such as Amazon for IaaS & STaaS, Google, SalesForce.com and others for PaaS & SaaS now enable “internal IT customers” such as entire business units or their individual end-users to fire IT with a credit card with instant gratification.

This current economic climate catalyzing the Cloud Computing trend represents an excellent opportunity for a strategic corporate “gut check” to re-evaluate core competencies and resulting areas of deliberate investment.  External Clouds (both Public and Private) represent a logical solution to the growing category of Tier 2 Enterprise Apps which no longer need to be provided by Internal IT departments.  Most of those apps can now be readily provided on a purely consumption basis via reputable Cloud Service Providers.

Continuing to deplete precious capital resources on non-differentiated services (i.e. Web Hosting, Payment Processing, Email, etc…) when they can realistically be supported via positive cashflow-funded operational budgets (paid purely on consumption to Cloud Service Providers) will diminish an organization’s competitive strength compared to more progressive industry peers who intelligently leverage trusted and commonly available Cloud services.

 

#7 - Avoid the “Private Cloud Trap”!

Pure-play, focused Cloud Service Providers are applying operational excellence and economies of scale to drive down the cost of computing services for commodity applications.  Most notably, Cloud Service Providers rigorously chargeback their subscribers for all resources consumed.

Many Private Cloud projects today (particularly new Internal Cloud projects owned by IT) over-emphasize new technologies at the expense of the overall contemporary corporate goals such as up-front cost-savings and long-term return on assets.

Therefore formal chargeback procedures (via internal funds transfer or proper external billing) are essential to successful Cloud projects.  In the truest capitalist sense, chargeback for resources consumed is the only way to drive the P&L discipline and efficiency priorities into your Cloud project for it to become and remain competitive with commonly available alternatives such as Amazon, Google, SalesForce.com and the like.

 

#8 – Hypervisor hedging

Thanks to it’s multi-year head-start combined with great vision and execution, VMware continues to dominate the hypervisor marketshare for Enterprise virtualization projects.  Strong Storage, Infrastructure, Platform and Software automation initiatives combined with rigorous chargeback for resources consumed can uplift common virtualization projects into the realm of true Cloud services.

However it’s worth noting that the hypervisor of choice for most of the successful highly-scalable Public Cloud services today are based on pure open-source hypervisor technology such as Xen and KVM on Linux or xVM from OpenSolaris, etc…

Finally, Microsoft is hard-core about regaining and retaining control of the primary abstraction layer between hardware and software for virtualization.  Their highly-anticipated R2 launch of HYPER-V is already predicted to re-balance the virtualization landscape for Enterprise hypervisors.

In turn, NetApp has been strategically prioritizing and balancing our R&D, certification and sales investments for a multi-vendor hypervisor landscape.  Much like competing against your best customers, or releasing yet another silo’d tactical product into a new market, placing all your eggs in one hypervisor basket is a very short-sighted approach to a long-term market trend.

NetApp has already been recognized for our wisdom in this area.

 

#9 – Invest in the right Vision

Cisco’s Unified Computing System has captivated Enterprise IT architects’ as well as Service Provider technical architects’ imaginations.  NetApp anticipated the same trends behind Cisco’s vision for the future of computing and made similarly aligned investments early on in our core technology.

Unified Computing requires Unified Storage.  Truly Unified Storage is a company-wide strategy, not a fragmented tactical product silo at one edge of a widely disjointed product spectrum.

 

#10 – Build on a strong foundation

Cloud Computing is founded upon the principle of infrastructure, platform and software repeatability via automation.  Spreading an infrastructure, platform or software grid too thinly across multiple architectures will directly decrease necessary economies of scale and operational efficiency, while increasing overall risk.

Truly Unified Storage is delivered via a cohesive product family, which encompasses a wide range of scalability, based on a single, consistent core storage architecture.  Data ONTAP is the strong foundation upon which NetApp’s Dynamic Data Center offering is based.  No other storage vendor offers anywhere near the same level of scalability (both up and down the product line), breadth of functionality and native multi-protocol support.

There are 5 pillars built upon a strong storage foundation which underpin a successful Cloud Infrastructure, Platform or Software Service.

Secure Multi-Tenancy:  Being the very antithesis of legacy dedicated physical infrastructures, Cloud infrastructures rely on the amortization of many subscriber services across as few physical assets as possible.  Resource sharing is required to achieve the economies of scale necessary for profitable delivery of Cloud services.  Yet Cloud subscriber expectations for data privacy and security remain the same as in the Physical world.  Imagine NetApp and EMC using the same Cloud service (as we actually do with TypePad’s blogging service).  We demand and expect complete isolation of our environments – particularly if stored on the same physical system.  Fortunately, NetApp, VMware and Cisco have been collectively working on a joint Cloud solution to this key requirement for quite some time.  Expect more details at our upcoming NetApp Cloud Launch later this year.

Service Automation & Management:  Much like resource sharing, Cloud Service Providers depend on operational efficiency to drive profitability in a highly-competitive low margin commoditized business.  Storage Architects have long admired the elegance of NetApp’s provisioning capabilities as well as our self-tuning dynamically optimized I/O architecture and simple management API’s.  Yet $/GB continues to be a key metric for Enterprise storage purchases.  OTOH - Cloud Service Providers depend on highly automated operating environments.  NetApp’s elegance is “monetizable” in this market, and is consequently valued as highly as $/GB by Cloud Service Providers.

Continuous Operations:  Unplanned outages are the bane of Cloud Providers existences.  In addition to immediate revenue losses, the damage to a Provider’s reputation and resulting sales pipeline is often far greater.  Harsh IT realities such as hardware firmware upgrades, software bugfixes and security patches as well as network updates often intrude on the high uptime expected by Cloud subscribers.  Accounting realities also determine the schedules by which equipment needs to be retired in favor of fresh gear.  The only way to maintain highly available levels of service is to have a storage foundation which enables the non-disruptive movement of subscriber data sets and workloads between systems.  This enables regular preventative maintenance, as well as load-balancing and disaster recovery capabilities key to continuous operation.

Extreme Storage Efficiency:  Object-level data protection has captured a lot of mindshare in the Cloud world.  Yet a 4-to-6 copy model (primary, local mirror, dr+remote mirror, backup & archive) is far from efficient at petabyte scale – despite any up-front capital cost savings on internal DAS storage.  It's no wonder the likes of Google & Amazon have to keep building giant data centers on a regular basis!

Legacy Enterprise Storage technology based on small RAID5 stripes or RAID10, thickly provisioned LUN’s and full data copies for scalability is an equally inefficient method of satisfying Cloud Storage requirements.  Fortunately, NetApp’s compounded and multi-layered storage efficiency technologies routinely offer over 90% physical storage efficiency and 600-700% data efficiency to Cloud Service Providers.  See Oracle’s multi-petabyte results above as but one example.

Integrated Data Protection: This is often an afterthought in most Cloud deployments today.  Not only does it contribute to the ridiculous 4-to-6 copy object model described above, but it also directly impacts SLA’s around data recovery.  Rich Data Protection (Backup, DR, Archive) solutions based on NetApp’s legendary scalable snapshots enable Cloud Service Providers to offer rapid RTO’s and abundant, granular RPO’s to their subscribers – potentially at higher margins or simply to gain marketshare.

 

#11 - Don’t rush to the destination

Back during the Y2K mania, NetApp began to enable the Storage Service Provider (SSP) Market via technology innovation such as our MultiStore capability (aka “Storage Hypervisor”).  Based on those early investments and key Cloud customer wins, NetApp is well-positioning to maintain our lead and capture more marketshare for Cloud Infrastructure and Storage requirements.

For most of 2009, Industry Analysts, Media & Bloggers have been gorging on the topic of Cloud Computing – and I am obviously now joining this herd.  Despite long-standing multi-year NetApp Cloud references like Yahoo! (SaaS Email), Oracle (PaaS “On Demand”) or Telstra (IaaS “OmniPresence”) it’s important to put the rapid evolution of this market into perspective.  It has a long way to go before it stabilizes and we all have much to learn along the way!

Comments

[Nigel]

I also believe this is the beginning of a new era in purchasing behaviour. Gone are the days when inefficient IT organizational structures enable stratgic decisions to be closed on a golf course with entitled sales reps from incumbent vendors!
In the new world of commoditized infrasrtcutures, real-world staffing and kit efficiency at scale will make or break Cloud provider business models.
This is an important early advantage for NetApp in the provider enablement marketplace, which I hope you fully capitalize on!

[Larry]

I think NetApp storage efficiency at PetaByte Cloud Scale could be the next killer app! Got any examples to share?

[J.F. Zarama]

Val:

great to see the evolution of NetApp towards facilitating storage on-demand for conventional infrastructures and for the clouds.

Your blog, as expected, is very informative. However, it is hidden and not easily visible to a casual end-user browsing www.netapp.com/us/

Given the strategic value of cloud services for the industry, indeed for NetApp, NetApp's cloud strategy and value proposition should be visible on the home page; Dave's blog and book reference, what a title, are also not easily found; they should, along with your blog and that of other colleagues; these are not personal information items but strategic statements reflecting NetApp's direction and hence they should be easily accessible from the corporate home page.

N.B. I found the title of Dave Hitz book interesting and reminded me of a saying: "the castrate the bull only once"

keep in touch;

jf.zarama at gmail dot com