When I read the announcement that IBM and VMware have signed a strategic partnership, that will in their words “accelerate enterprise hybrid cloud adoption”, it came as somewhat of a surprise. Not an earth shattering "they've figured out how to commercialise data storage using DNA strands!" surprise, but still a surprise.
Essentially, what has been announced is a solution to extend your on-premises SDDC (Software Defined Data Centre) built on VMware technologies (vSphere, NSX and Virtual SAN) into the IBM cloud based on IBM SoftLayer BMaaS (Bare Metal as a Service), where pre-configured VMware SDDC environments get automatically provisioned.
Like most pieces of surprising news it elicits several different reactions depending on your perspective. The news also reminds me of some of those 2016 predictions of late last year, that we will see the end of enterprise cloud (which are proving to be true when we think of the recent formal news from Verizon). This partnership between IBM and VMware is potentially the morphing of another enterprise cloud into something likely more relevant and more accessible.
A cloud on-ramp for those wary of cloud
There are the cloud embracers who got on board early or even just now at what we can still call the early majority. However, there are those who have significant investments in legacy infrastructures who have not yet gotten on board the cloud train. For those that are disinclined to move to cloud, they have many reasons, which each individually might not appear as show stoppers, but that can easily be stacked up like dominoes. And it can be difficult for cloud advocates within those organisations to develop any momentum.
This announcement taps into that opposing or cautious sentiment, with a safe on-ramp to cloud for those organisations with significant legacy investments. It will likely also appeal to those who are concerned that a move to public cloud is too hard because it requires a re-architecture of applications and a rethinking of traditional concepts and approaches to availability, data protection and finances. Which, by the way, is what the cloud advocates love about cloud.
The problem with credibility
Part of the challenge this partnership will have is with the credibility of the strategy from these two major players. VMware have come late to the public cloud game with vCloud Air, and there have been a few shifts and developments in direction already; with Google partnerships announced a year ago that still show a lot of promise but not a lot of action yet; with planned partnerships with Virtustream that they then walked away from when it turned out to be the albatross around the neck of their share price; and now this partnership with IBM and SoftLayer data centres. This makes credibility somewhat of a challenge, however, execution and delivering results are the perfect panacea for that. I suspect that if we see some reasonably quick successes, then past vacillations in strategy will be forgiven just as quickly in this cloud eat cloud world we live in.
It kinda makes sense
Several years ago, VMware realised that they needed to overcome barriers of perceived complexity of the private cloud stack. Complexity, like everything in this world is relative of course. OpenStack is notorious for it, and the common sentiment seems to be that it works great, so long as you don't want to upgrade - at which point the interoperability complexities will drag you down into the quicksand. For VMware, EVO:RAIL showed great promise as a way to rapidly deploy and deliver (through a variety of hardware partners) on the value of SDDC (everything virtualised, centrally controlled and orchestrated/automated), however the uptake was poor. This was probably for a number of reasons, including complexity of the positioning of the offering and cost barriers when compared to roll-your-own approaches. So, now we at ViFX are focussed efforts on how to quickly deploy and deliver on the value of SDDC, and VMware are also looking to refresh EVO type approaches.
From the IBM side, this partnership looks to be tapping into the same concept of rapid delivery of SDDC value by automating the deployment of the SDDC stack (onto bare metal). In a crowded market, IBM carved out a small niche with their unique approach (through the acquisition of SoftLayer) using bare metal to solve the challenge of rapidly deploying configurations on that bare metal (OSes, VDI or complete Virtual Infrastructure). And IBM is also rightly getting some credit for not playing to type and hugging the tin, but rather letting go of the hardware and realising the value is all in the software, particularly with their solid and emerging offerings in the software defined storage space (and others through acquisition of early mover object storage specialists Cleversafe).
So, thinking this announcement through and seeing that IBM and VMware have this same approach of looking to deliver on the value of the SDDC stack, through rapid automated deployment onto bare metal; it all kinda makes sense.
Trying new things
So, will the market give both of these vendors another chance? Do two chancers make a romance-a? VMware can be accused of changing tack again. IBM are looking to overcome accusations of an inability to execute (especially poignant in New Zealand when you consider previous local cloud attempts and the current beleaguered health infrastructure platform).
The thing that strikes me about our industry is that there is an overwhelming driver to call and pick the winners and dodge the losers. But of course it’s never that clear cut. So what’s an IT pro to do? We ourselves have been advocating for a while that cloud/rental/OPEX models mean that traditional approaches to slowly deliberated, high spend, strategic investments in enterprise software are no longer desired or appropriate.
Research it, poke it, try it, rent it, use it – and if it doesn’t work, chuck it and go for something else. That’s a bit flip, but the idea is that experiments aren’t failed. If you out-grow a solution or something better emerges, go and take advantage of that new thing. If you think you want to do cost chargeback, start with a spreadsheet, then try a simple SaaS tool, and then maybe go for something more enterprise once you’ve overcome those operational hurdles. etc. etc.
So maybe we can argue the same approach is valid at a multi-national organisational level. VMware have tried a few things. IBM have tried a few things. But this could be the one. I wouldn’t put big bets on it, or go all in, but I certainly wouldn’t turn my back on it too early either.
This is of course a very quick analysis and there are always a number of angles to come at it from. I'm sure we will continue to hear more from both IBM and VMware, and we'll gradually see how this plays out in the market. As for a ViFX public stance or views to partnership? We will always want to lift the lid and understand something better. And we'll always want to share our views with our customers and consider how it fits with their business roadmap and IT strategies.