Nirvanix Fallout re: Cloud Market

Brian Wood Blog

The unexpected exit of fellow San Diego firm Nirvanix from the cloud storage playing field is analyzed in the piece below from 451 Research analysts Carl Brooks and Simon Robinson.

AIS customers and prospects can take comfort in knowing that we have been in business for over 20 years and we have very strong financial backing from our three committed private equity investors.

In other words, we are firmly established and financially sound (read: profitable).

So relax — we aren’t going anywhere!

Emphasis in red added by me.

Brian Wood, VP Marketing


Nirvanix’s exit: the impact on the cloud market

The sudden announcement by longtime cloud-storage provider Nirvanix that it will be shutting down at the end of September, and the subsequent scramble by partners and users to recover data before the lights go off is remarkable in several ways. Our previous spotlight focused on what we believe are the primary reasons for Nirvanix’s failure; here we look at the wider implications of its failure on the overall cloud model and market.

First, the scale of the shutdown is large. Nirvanix had about 40 petabytes of unique customer data under load – spanning well over 1,000 customers – all governed by SLAs and consumed around the world. That puts Nirvanix in the company of the largest consumer backup providers, such as EMC’s Mozy service, although not on the scale of Amazon Web Services or Google or Microsoft. That is a staggering amount of data to think about moving in less than two weeks over commercial ISP trunks, especially when thousands of users are going to be attempting it simultaneously.

Second, the alacrity of the stoppage is unusual; even in dire straits, a provider can usually find a way to shut up shop and then gracefully retire accounts and allow data to be moved. According to reports, partners and customers were simply notified directly that the end of the month was the end of Nirvanix, although it seems work is under way to extend this into the month of October.

But at least one partner, Aorta Group – which runs cloud storage business Aorta Cloud, and private capital fund Aorta Capital – has announced an effort to raise cash and do something with the Nirvanix business. Aorta Capital describes itself as a ‘specialist in merchant banking, turnaround financing, restructuring, project financing and asset management.’ Nirvanix had about $70m invested in it over three rounds of fundraising, and has seven datacenters.

Aorta has floated a few options to secure Nirvanix’s data and some of its assets, which include condensing active sites to two or three datacenters (data-replication practices mean that this is not as daunting a task as it might seem), a possible acquisition and transfer of assets to IBM (IBM Global Services is a ‘strategic partner’ of Nirvanix), the involvement of another UK-managed infrastructure provider (BroadCloud) and similar approaches. Aorta says via a public notice that IBM is committed, a bank has been found to match any funds raised, and that as much as $1m in rescue cash is currently available with more to come. None of this is a done deal – and IBM has also indicated that it is working to transfer Nirvanix customers to its recently acquired SoftLayer property – but it appears to have some validity.

Potential impact: sentiment and risk

It is unlikely that Nirvanix will be able to continue normal operations no matter what. The impact of its abrupt action was felt immediately by its network of partners, including many multi-tenant datacenters and service providers like DRFortress, CoreSite, CyrusOne, Symantec, CA Technologies, IBM, Dell and many others. The larger tech-centric firms are going to be unhappy, but the MTDC providers, IT services providers and resellers will likely take severe blows to their reputations and lose the confidence of customers.

In many cases, these providers have brought in a service like Nirvanix and white-labeled it or offered SLAs on top of the service. They will take the blame for lost data, or the sudden disruption and expense in moving data, and they will suffer for having chosen an unreliable partner in Nirvanix. This has happened before – Iron Mountain and EMC both shuttered cloud storage services driven by partner sales, albeit on a much smaller scale than Nirvanix. In those cases, however, there was a year or more of time to transition, not two weeks.

In the longer term, the impact will be defined by the extent of the failure. If Nirvanix goes dark and data is actually lost, it will be seen as a definitive black mark for the cloud market overall. This kind of situation is the basis of nearly every ‘anti-cloud’ sentiment for enterprise IT organizations – no matter how ‘cool’ or inexpensive, the cloud is simply untrustworthy. The anti-cloud revanchists in the enterprise now have serious ammunition to argue against using external providers. An event like this does far more to retard adoption in real terms than any temporary outage. IT people understand that technology can fail, yet they are not so understanding when they feel their trust has been abused, especially as private cloud products are flourishing.

Likewise, partners that get burned by this are going to be far less willing to experiment with a smaller vendor or a startup service provider. Instead, they will turn to traditional vendors or larger vendors, even if there’s better technology available. That rare bird, the infrastructure service provider startup, will also face a higher bar. Seventy million dollars is a lot of money to throw down a hole, and this will likely influence future investment decisions for a bit. Fortunately, Silicon Valley doesn’t have a great short-term memory in a booming market.

Competitors step in

Again, fortunately, the managed infrastructure and cloud markets are booming, so this isn’t a hole at the waterline. Competitors like TwinStrata, Egnyte, Global Net Access (GNAX), Savvis, IBM and others are scrambling to nab Nirvanix customers. The storage services market encompasses practically every hosting provider of significant size, as well as a host of SMB and consumer-facing providers: The Go Daddy Group, EMC/Mozy, Barracuda Networks, Comodo Group, Box, Dropbox, Google, Microsoft and others. Nirvanix didn’t have any capabilities that others can’t essentially match in different ways, and the appetite for external resources is far larger than the number of disgruntled Nirvanix customers will be.

The 451 take

Running IaaS should entail minimal risk at every turn, from cheap gear and low to nonexistent marketing and support costs, to being able to build only as demand grows and get into the black as fast as possible. It’s a business first and a gamble last, and Nirvanix didn’t pull it off. The worst outcome is that there is a chilling effect on investment cash and the emerging cloud channel, and innovation suffers; but we feel that the impact will be short term and unsustained. It’s a definite boost for the giants, which become the default when enterprises don’t want to take a risk.