If you had asked me, a year or so ago, which of the three largest public cloud vendors was most likely to offer a compelling hybrid cloud offering, I would have said Microsoft without a doubt. After all, AWS is vehemently adamant that the public cloud is the right way forward, and that anything less than that is highly sub-optimal. Microsoft, on the other hand, has an extensive pedigree as an on-premises vendor, has been talking about hybrid cloud (via a couple of different initiatives) for the longest time and arguably has the most extensive footprint of traditional enterprise customers – those most likely to demand a hybrid offering.
That assessment would have almost entirely overlooked Google. After all, of the three companies, Google is, to its core, the firmest demonstrator of a public cloud approach, it has also had the tendency, at least in the past, of dismissing approaches that didn’t align with its worldview. Surely a company like Google, with its purist, technology-led approach, would be the last one to embrace hybrid, and hence private, cloud approaches?
But before we look to what has changed, it is worth noting that a couple of years ago Google appointed Diane Greene, co-founder of VMware and a hugely respected powerhouse in enterprise technology circles, as the CEO of its cloud division. Greene quickly set about changing the organization from the top down and installing the sort of robust enterprise-focused things that the company needed – support, professional services, a partnership strategy – all things that Greene has put in place.
And so, perhaps it is less surprising than we would have initially thought that at this year’s Google NEXT conference, the company announced a number of developments that see it firmly plant a stake in the ground and announce its intention to become a best-of-breed hybrid cloud vendor.
The announcements
The announcements that, in my view at least, are most interesting, impactful and, perhaps surprising, are the ones which take Google away from its traditional approach (both technological and commercial) and move into the realm of a broader enterprise technology vendor. Some highlights:
-
Serious investment into the partner channel. I’ll talk about these aspects in another post but for now, suffice it to say that Google is spending up large in enterprise sales and support, professional services and developing its global channel partners. Particularly surprising was Greene’s stated ambition that 100% of deals for Google Cloud will have external partners attached to them
-
GKE On-Prem is an on-premises version of Googles existing Kubernetes Engine product but, in this case, and as the name suggests, it is fully managed by Google for deployments across enterprise data centers. The service includes access to all of the standard features of GKE, including the ability to register and manage clusters and monitor them with Stackdriver, as well as identity and access management. It also includes a connection to the GCP Marketplace, which recently launched support for Kubernetes-based applications. Using the GCP console, customers can seamlessly manage all their GKE clusters, be they on-premises or in the cloud. The oft-mentioned single pane of glass vision is realized. It should be noted that GKE On-Prem will be available in alpha later this year with an unknown GA date at this stage
-
Knative, a new project that aims to help customers set up serverless infrastructure (and, yes, that is something of an oxymoron) that runs the same code across multiple public clouds or on-premises. Built on top of Kubernetes alongside the Istio service mesh, Knative aims to abstract away all the requirements for preparing and running an application in a cloud-native way
-
A new CI/CD offering, Google Cloud Build, which is deeply integrated with GitHub (interestingly enough given the recent acquisition of that company by Microsoft)
The competitive position
While these announcements were, as I already mentioned, something of a surprise from Google, they don’t exactly come out of the blue from the broader cloud industry. AWS has a dual-sided approach to hybrid. Firstly via its edge devices and secondly via a partnership with VMware which is still in its earliest stages.
For its part Microsoft has long talked a very strong hybrid cloud story but has been lacking when it comes to execution – Azure Stack, which was finally released a few months ago, has been a very, very long time coming. There is also some confusion from customers and partners as to what Azure Stack actually is – many still think it’s a “cloud in a box” and hence a great way to avoid the continued march to the public cloud. Microsoft doesn’t see it this way but the confusion still exists.
Of the three big vendors, Microsoft is, as I’ve already mentioned, the most natural hybrid cloud vendor and it has been a frustration to many cloud pundits that Redmond has taken so long to get to this point. That said, Azure Stack is now out into the wild and one would think that its availability, as well as the breadth of services that are supported on Azure Stack, will grow rapidly.
The big unknown, as is often the case, is AWS and its intentions. We’re only a few months away from re:Invent, AWS’ annual conference and I would be very surprised if we didn’t see some very strong moves in the direction of hybrid from the company. AWS already has Snowball, its mass-storage device designed for on-premises implementations, I’d not be at all surprised to see compute offered from Snowball and, if AWS went down this road, that would be essentially a fully hybrid AWS cloud service available on an already in-market platform.
MyPOV
It needs to be stated clearly that these on-prem offerings while planting a huge stake in the ground, aren’t generally available yet. As one of the attendees at the analyst summit opined via Twitter, an announcement, and general availability, with all the enterprise expectations that go with it, are two different things:
GKE On-Prem is in alpha until later this year, then beta until maybe 2019 sometime? It will be more interesting after it has an SLA. Demonstrating multi-cloud interoperability to several nines is apparently not as easy as talking about it…
— Paul Teich (@paulrteich) July 24, 2018
That said, one shouldn’t understate just how much of a stake in the ground Green and her troops are planting. The fact that Greene’s #2 and the man credited as the technological brains behind Google’s cloud initiatives, Urs Holze was so bullish as to state unequivocally that hybrid is a reality and the normal case for enterprises says it all. He’s also more pragmatic than I would have expected given his Google pedigree and he expressed a real desire to make the two worlds – that of the cloud and on-premises – look “more similar rather than more different.” If only that was a standard approach from cloud vendors!
I came away from NEXT generally impressed by what Green has put into place over the past couple of years and bullish about Google’s opportunity. Sure they’re a distant third in the cloud race (a term that Green reportedly detests) but they’re making some good moves that, all being equal, should help them compete with those two Seattle companies.
