Google today has announced it has acquired VMware-as-a-Service provider CloudSimple for an undisclosed sum.
This acquisition has some interesting competitive dimensions, particularly relating to Microsoft Azure. CloudSimple is currently the main mechanism for running VMware on Azure and GCP, while VMware on AWS is jointly engineered between VMware and AWS. It’s a more tightly integrated offering, and not subject to this kind of pull-the-rug-out-from-underneath-us manoeuvre.
There is a Virtustream-based mechanism for running VMware on Azure in the works, but it’s not available yet. It’s been advertised as being available in late-2019 but given we’re in November already, there’s not a lot of 2019 left.
A Microsoft spokesperson commented via email that “Azure VMware Solution is a Microsoft Azure offering that is supported by Microsoft partnering closely with VMware. Microsoft and VMware are continuing to invest in Azure VMware Solutions to enable new scenarios, for example, new HCX capabilities making migration even simpler for our customers. There are no changes at this time.”
I approached Google, who referred me to their own blog post, and that of CloudSimple founder and CEO Guru Pangal, but stated that Google would not be making any further comment. It’s important to note that Google has already been using CloudSimple to provide a way to run VMware on its cloud since July of 2019.
By acquiring CloudSimple, Google has turned Microsoft into a customer as well as a competitor (and also given it a payout as an early investor in CloudSimple), but I consider it unlikely that Microsoft will want to continue using CloudSimple to provide its VMware-on-Azure offering, given the competitive tensions of public cloud. The pressure to bring the Virtustream offering online has just been increased substantially.
But given we’re talking about public cloud, why does it matter that Google has acquired a way to run on-site workloads in its cloud? Why would anyone care?
The answer is enterprise.
Enterprises love VMware, and have very large install bases of important applications running on it. They’ve also invested a substantial amount in the tooling and expertise required to operate VMware as critical infrastructure. Moving these workloads to public cloud without VMware would involve fundamentally re-architecting the applications in most cases, and in all of them would require learning how to operate them in a completely new way.
While we can argue about whether that’s a good idea in the long term, we have to face the reality that enterprises don’t move quickly, and that this is a multi-year—and potentially multi-decade—proposition. An option to add public-cloud flexibility but retain most (if not all) of the operational setup I already have is compelling to an enterprise that doesn’t want to change everything overnight.
We’re currently seeing a transition from the early days of “public cloud is the One True Way” to a more pragmatic approach where multiple options coexist. The driver for this transition is the huge pile of money that enterprises spend on IT each year.
Of all the public clouds, Google was in the worst position with providing options to enterprise before this acquisition. AWS is still the market leader, but Azure is rapidly catching up, and Microsoft is a trusted brand with the enterprise in ways that AWS hasn’t managed to crack (yet). Google was a distant third, but by making it easier to run VMware on Google Cloud, and by snatching the Azure method out from under Microsoft’s very nose, Google has bought itself a chance at some decent enterprise market share.
This is an opportunity for Google, but it is in no way a done deal. There are still many hurdles to convincing enterprises that VMware-on-Google is the right choice, particularly given Google’s penchant for killing off products that people like using. While AWS and Azure are substantial competitors, Google’s biggest problem is Google.
Hopefully with the CloudSimple acquisition Google can start to get out of its own way when it comes to enterprise business.Tags: cloud, cloudsimple, gcp, google