Back in 2003 hardly anyone had heard of a small but rapidly growing technology company called VMware. But since that time VMware literally exploded to become the dominant player in enterprise server virtualization. Now the same forces that propelled VMware to the top of the server virtualization heap are at work again. Only this time the forces are coalescing around a cloud storage company called Nirvanix that could and should result in it riding the same type of wave that carried VMware to market dominance.
To understand why Nirvanix is in the same type of enviable position that VMware was in just a few years ago, one first needs to look at the combination of factors that contributed to VMware’s dramatic rise from obscurity to the top of the enterprise server virtualization food chain. While in most cases these factors were arguably outside of VMware’s control, by VMware’s proper management and possible manipulation of these forces it was able to slingshot itself past much larger competition (Citrix, Microsoft, Red Hat) into the enviable position of market leadership that it holds today.
So at a high level, here are the four (4) phases that VMware went through that resulted in where it ended up today:
Phase 1: VMware went after the tier 3 and lower application servers that everyone was ignoring but were a huge administrative burden in enterprise organizations.
Tier 3 applications consist of file, print and web servers while Tier 4 applications are primarily test and development servers. So what made and make these applications such a burden for enterprises to manage is that there may be tens, hundreds and, in extremely large organizations, possibly thousands of these types of application servers.
But the real problem from an enterprise server is not necessarily the number of servers. If there is revenue attached to them, organizations will find budget to manage them. Rather it was the lack of budget and IT staff time required to adequately support and maintain them.
In many cases, the hardware platforms on which these applications landed was hardware that previously hosted business critical applications but that was now either out-of-warranty or no longer had sufficient performance or capacity to host these applications. So they were repurposed for these less critical apps.
On the surface this makes sense. But older hardware is more prone to fail and, because of its age, more difficult to fix, takes longer to fix and requires more IT staff time to be spent on doing break-fix tasks for non-critical application servers. Aggravating the problem is the fact that each of these physical servers consumes power and puts out heat which runs up operational costs.
In response to these problems VMware announced, “We have a solution!” VMware told organizations that they could take 5, 10, 20 (or whatever) number of physical application servers, virtualize them and put them on a single physical machine.
Better yet, at least from the perspective of the application owners and the IT staff supporting it, these single physical machines were new servers. They could now get rid of their old servers that are a pain to maintain and cost justify the expenditure on the power savings associated with eliminating the 20 (or whatever) servers while gaining new levels of flexibility in management and increased application performance.
But maybe what was the most important item to VMware succeeding in this Phase 1 is that it accomplished this task without upsetting the other enterprise OS players like IBM, HP, Microsoft, Red Hat, and Sun. As a result, it was given largely free reign in an area that was a huge pain point in companies and resulted in VMware capturing a sizeable and growing portion of the data center.
Phase 2: VMware gained momentum by partnering with the major players with a “Win-Win” strategy.
VMware virtualizing Tier 3 and Tier 4 application servers did not only solve customer problems. It solved a number of problems for OS, server and storage players as well and created new opportunities for them.
- Dell, HP and IBM loved that VMware provided a built an argument for enterprises to justify ridding themselves of all of those troublesome out-of-warranty servers and replace them with new server hardware.
- Storage companies found out that as virtualized servers replaced physical ones enterprises needed new external storage for these virtualized machine. So they were now selling storage into a part of the enterprise that had rarely, if ever, needed external storage. This made VMware like a double bonus for them.
- VMware virtualizes every OS – Linux, UNIX and Windows – so what made VMware especially appealing to OS vendors was that in test and dev environments, VMs have this nasty habit of growing like jack rabbits. So every new VM meant a new licensing opportunity for them.
So with everyone happy, they all became willing VMware partners and sought to introduce VMware into every account that would to listen to the VMware story because it was a win-win-win for everyone involved. As a result, VMware got more traction, OS, server and storage vendors sold more and users lowered their operational costs.
Phase 3 – The VMware community, knowledge base and level of exposure grew such that VMware started to be used to virtualize tier 1 and tier 2 applications
As VMware became more pervasive, enterprises became more knowledgeable and adept in managing it. Further, the early problems and objections that occurred among tier 3 and 4 applications were addressed and overcome. It finally got to the point that organizations recognized they could confidently and safely start virtualizing Tier 2 and higher applications. This was the tipping point where VMware went from a handy technology in test and dev environments to a viable enterprise platform.
Phase 4 – EMC buys VMware and further legitimizes VMware’s position as an enterprise ready platform by providing the enterprise level of support VMware needed.
Once VMware began to virtualize Tier 2 applications, EMC pounced on it. It had already seen the evidence of the uptick in storage consumption that resulted when Tier 3 and 4 apps were virtualized but with VMware tackling Tier 2 apps, VMware needed someone to come alongside it to provide the level of credibility and support that enterprises expect and want. But equally important, EMC continued to be give it the freedom to develop which further contributed to VMware’s rise to fame.
Now granted, other features and technologies emerged in VMware along the way that accompanied and facilitated its growth in each phase that led to it transitioning from an interesting start-up to an enterprise platform. Clearly features like Site Recovery Manager, the vStorage APIs and many others helped pave the way for its transition from one phase to the next.
But the introduction of the features it needed to succeed at each stage was predicated upon successes it had already achieved in the prior phase. Further, by not trying to jump immediately to phase 4 and instead starting at phase 1, it worked its way slowly up the enterprise stack and proved itself.
In the process, VMware cut its teeth on areas within enterprises that had low visibility, did not put organizations or their business processes at risk and solved real business problems. But most importantly, VMware got to understand how enterprises were using its software even as they were serving as the proving ground for its software leading to what we have today: the first enterprise-ready server virtualization platform that facilitated the introduction of a cost-effective cloud computing solution into almost an
y size organization.
But what I believe is still one of the inhibitors to cloud computing becoming a true reality and achieving wide spread adoption is the introduction of a corresponding cloud storage platform that has the same ubiquitous nature of VMware. Because until one decouples storage virtualization from the underlying storage hardware and storage software in the same way that VMware decoupled server hardware and server OSes, cloud computing will still be impossible to achieve in terms of massive adoption.
But when one looks at these phases that VMware went through on its path to success and then compares that to the storage companies that are attempting to achieve something comparable in the storage space, there is only storage provider that so far meets all of these criteria and that company is Nirvanix.
Consider where Nirvanix is at in regards to the four phases discussed above:
Phase 1: Like VMware in its phase 1, Nirvanix is going after all of the problematic data (archive, backup, unstructured data) that many large enterprise-class companies do not have the time or budget to justify managing.
Archive, backup and unstructured data that now resides within organizations share some of the same characteristics that file, print, web, test and dev servers possessed before they were virtualized. Some of this data resides either on old storage that is no longer useful in production or the cheapest most reliable storage that companies feel comfortable procuring.
In either case, organizations are not apt to want to spend one second longer than they have to maintaining or managing this storage or the data on it. The only time they really care about this storage is if it fails (they have to fix it) if someone needs to access data stored there (a recovery, an eDiscovery, a random request) or if there is a disaster where they have to get the data back. In other words, organizations understand this data has value, it is just that its value is difficult to quantify so the budget dollars and resources needed to manage it are scarce.
This is where Nirvanix plays. It provides an enterprise cloud storage solution that enables organizations to store data in such a manner that it is readily accessible, is cost effective, is stable, is reliable, is scalable, provides adequate performance and solves the problems that they most need to solve right now.
Further, Nirvanix is “pay for play.” In other words, enterprises only pay for the amount of capacity they use. If they use more, Nirvanix charges more. If they use less, Nirvanix charges less. But regardless of what they pay, enterprises do not sacrifice on availability, reliability, uptime and scalability. Nirvanix manages and/or provides all of that in the background.
Now what users do not get right now from Nirvanix is adequate performance or throughput to satisfy the needs of tier 1 or 2 applications, but guess what? Nirvanix does not care about that (at least at this point) and is not currently targeting that segment of the storage space. Nirvanix is leaving that to the storage vendors with high-performance systems designed for Oracle databases and ultra low-latency transaction processing like EMC DMX/VMax, Hitachi VSP and HP 3PAR, among others.
All Nirvanix does is provide a cloud storage infrastructure in which enterprises can economically and confidently store data that is of uncertain value to organizations and then leave it up to Nirvanix to do the rest. This includes managing the back end storage infrastructure, handling the replication, handling the technology refreshes, making sure the right data is on the right storage, everything. Further, Nirvanix delivers it in public, hybrid or private cloud configurations–all with its “pay for play” model.
This is incredibly appealing to organizations and explains why customers were standing 2 and 3 deep at its booth at the recent NAB and Symantec Vision shows and why Nirvanix was literally signing deals on the show floor. In fact, while I was trying to talk to its CEO at Symantec Vision earlier this week, he had to break off our conversation to engage a new customer who wanted to talk about moving multiple petabytes of data onto Nirvanix’s cloud. So phase 1 of Nirvanix becoming like VMware is already complete.
Phase 2 – Like VMware, Nirvanix is gaining momentum as it is aggressively being sought out by other major OS, server and storage vendors in the market.
Nirvanix does not care which applications or server operating systems send data to it or which storage it virtualizes. While a number of storage platforms can claim the former, very few or none can claim the latter. Unlike most other cloud storage providers, Nirvanix cloud storage software can layer atop any storage hardware platform from any vendor and any file system and it also has CloudNAS software that can be deployed as a virtual appliance.
Its cloud software IP is completely hardware-agnostic and file-system agnostic–it can integrate on whatever you throw at it. Nirvanix can basically take its own file virtualization layer and put anybody’s storage underneath it and present all of this pooled storage as one large global namespace.
While “global namespace” is used in multiple ways, Nirvanix represents it as a virtualized layer that sits on top of the actual files, so that no matter where you upload or download a file from, it always appears as the same file. On top of this Nirvanix then adds other cloud capabilities–like metered billing, support for billions of objects, secure multi-tenancy–to existing storage platforms, making it more valuable to customers in the process.
Now exactly what storage platforms Nirvanix has virtualized on the back end in the creation of its storage cloud, I do not know and do not know if it matters a great deal. But what is important is that if a customer wants to use Nirvanix as its cloud storage provider and it has a requirement to keep all of its data on EMC storage on the backend at Nirvanix’s site or in the customer’s own private data center, Nirvanix can do that by setting policies so that customer’s data is only stored on EMC storage.
Granted, the customer will probably pay a premium for that privilege but Nirvanix can accommodate that. Conversely if a customer says “just store its data on whatever” and tells Nirvanix to make sure that it can access the data locally or remotely, Nirvanix can facilitate those needs as well.
It is for reasons like these that enterprise application, OS, server and storage companies are actively seeking out and getting partnerships with Nirvanix. Already companies like CommVault, Dell, HP, Front Porch Digital, Symantec and many others have already created partnerships with Nirvanix even as some of these companies are simultaneously touting their own cloud solutions that could possibly be perceived as competitors to Nirvanix.
However Nirvanix does not see it that way. Because its solution sits above or below the respective offering of each of these companies, it really does not care if these other companies say they offer “cloud.” Nirvanix views their offerings as primarily “cloud-like” since Nirvanix either virtualizes them or acts as a cloud storage target for their solution.
Storage hardware providers can sell more product by making their systems cloud-enabled with Nirvanix software, be it in public, private or hybrid cloud storage configurations. Software providers can sell more licenses by helping their customers migrate data to the cloud by integrating Nirvanix as a cloud back-end that ships with their software products. So you essentially have a company that is providing the middle-glue-sticky-layer that keeps every side (its customers, partners and Nirvanix) happy as more products are sold and customer problems are solved. This again coincidentally sounds just like VMware.
So that e
xplains why Nirvanix is getting in a cozy relationship with so many of these providers. Much like what happened with VMware, both Nirvanix and its partners view the partnership as a “win-win” because they both get the cloud definition they want even as customers get a more efficient infrastructure.
In terms of phases 3 & 4 for Nirvanix, those have yet to occur but I would argue that Phase 3 is well underway. Nirvanix has a very referenceable and established customer base whose knowledge of its product and comfort level with it is expanding and growing. After all, it would not attract Tier 1 customers like Cisco, NBC Universal, Fox, Logitech, Johnson & Johnson and Comcast without having some secret sauce. So as its customer base grows, Nirvanix appears to be rapidly moving toward that magic tipping point where it virtualizes tier 1 and tier 2 storage and the applications associated with them.
It is for these reasons that I see Nirvanix poised to become the next VMware. To date its history and successes closely mirrors VMware with Nirvanix having essentially successfully completed phases 1 and 2 in this 4-phase journey to market dominance. So assuming Nirvanix executes on the next two key phases, it will have put in place the final pieces of the puzzle necessary to enable massive market adoption of the cloud. It is at that point that Nirvanix will emerge as a new major player in the cloud storage market who will be sharing the cloud spotlight and stage with VMware.
I have some further thoughts on this subject which I posted in an October 14, 2011, blog entry.