What Cisco does is always news, even when it involves what’s usually a dull-as-dishwater theme like reorganization. Cisco gave surprisingly negative guidance in its most recent earnings call, and everyone is thinking that the reorg has more to do with that than gaining “economies of scale” or “aligning with the evolving needs of customers”. I think they’re right.
Network vendors have always had a problem prying budget money from CIOs, but in the last 10 years the problem has gotten worse. On one hand, big IT vendors like IBM have uniformly lost strategic influence with customers; on the average their influence in 2019 was half what it was a decade earlier. On the other hand, IT and data center issues continue to rank higher than network issues in setting strategic priorities.
The cloud, which started to impact enterprise strategic plans in a meaningful way about five years ago, offered network vendors a bit of a respite. There was a general view that cloud computing was going to undermine if not displace the data center, and that view stopped the steady growth of data center influence that we’d seen, though it didn’t reverse it. But now enterprises are seeing the cloud differently, and that’s what I think is making Cisco see things differently.
We were never going to move everything to the cloud; that theme was click bait. The cloud is going to impact every enterprise, and every application in large enterprises, but it’s augmenting and not replacing the data center. Thus, the data center remains a power center, and that truth resulted, in 2019, in the first gain in data center influence on overall strategy since the cloud came along.
Strategic influence drives technology planning, and that drives long-term buying. If you just turned in unsavory financial guidance, you’re probably interested in long-term buying trends, and that’s why I think the Cisco moves are tied to these cloud/data-center dynamics trends. Organizational charts don’t sign checks or even drive product strategies, but they do make sales and product planning easier if they align you better with buyer thinking. Let’s consider Cisco’s moves in that light.
Cisco is combining its Enterprise Networking and Data Center Networking groups, and to me that’s a logical response to the fact that hybrid cloud reality is making the data center the focus of the network (again). For a decade, as I’ve already noted, the data center was the primary strategic focus of enterprise networking, and now that the cloud threat has lifted, it’s set to resume expanding its influence. From a sales perspective alone, that makes having a data-center-centric networking strategy essential, and that’s not likely to happen with the two kinds of networks served by different technology teams.
If that point doesn’t signal that this stuff is connected to you loudly enough, consider the other half of the announcement. Cisco is renaming (to Cloud Strategy and Compute) and changing its current cloud computing software unit to include data center software and servers. Think about it; if you were going to define a business unit to sell to hybrid cloud users, what would you want in it? Cloud software, data center software, and servers.
Still not convinced? Try this on: Dave Ward, a super-network guy who was Cisco’s CTO of engineering and chief architect, is being replaced by Roland Acra, who now heads (wait for it…) Cisco’s Data Center unit. Dave’s new role has not been announced, but it’s clear that at every level, the data center is taking over Cisco’s strategic focus. Router networks, once the dog in the analogy, have become the tail to data center networks.
Cisco has read the market tea leaves, and started to design teapots to fit, right? Not so fast, because the market has plenty of tea-leaf-reading firms, and many of them are seeing not something different, but the same thing in a different way. I think that the real story of the future hybrid cloud is being told by the transformation of Kubernetes.
Kubernetes was a Google development, something designed for application deployment and redeployment in the online OTT world of microservices and stateless, fully scalable, software. Just three years ago, it was an interesting if somewhat geeky story, and now it’s hard to pick up a tech rag (even a virtual one) without seeing Kubernetes somewhere on the lead pages. Kubernetes has grown from a kind of container DevOps to the centerpiece of an ecosystem.
An ecosystem of what? That’s the real transformation. Kubernetes and its ecosystem used to be focused on cloud-native applications, applications designed for the cloud and its elastic, agile, expansive vision of user clouds and resource pools. Now the Kubernetes ecosystem is heading…to the data center. The reason is that the future demands a hybrid application platform that may be driven by the data center, but has its head in the cloud. Kubernetes, by creating an application model that crosses that kind of complicated boundary, becomes the hot strategic property.
Cisco isn’t exactly a Kubernetes ecosystem household word. In fact, there are three players that are really driving the strategy change, and they are VMware, IBM/Red Hat, and Google, in order of influence. Notice that network vendors whose names start with “C” (just for example) don’t appear on this list. In fact, while Dell is surely linked with VMware, and while HPE (who’s often named in the media as the company Cisco is counterpunching with on this announcement) has also just made a Kubernetes announcement, it’s VMware carrying the water on this one for Dell, and HPE’s Kubernetes story is weak and behind the times. Thus, nothing in this Cisco tale seems linked to the real driver of change here.
Cisco needs to recall that every enterprise already has a data center. Nearly all have cloud strategies already, too. What most of them don’t have, and what they now need badly, is something that links the two into one IT strategy, into something that doesn’t turn into one of those multi-year-nobody-wins integration slogs. They also need to avoid their tendency to turn every announcement into meaningless high-level drivel (like their current one) or into unfathomable low-level mire and muck (their optical one). It’s like they have a war between their brain people and their mouth people, and only one can win. As a result, nobody ever wins, really.
There is a clear problem here, a clear need to define a hybrid cloud ecosystem. Kubernetes is almost surely going to be at the center of that, so there’s no point trying to differentiate there. Beyond Kubernetes, there’s plenty of room to build and grow, and that’s where Cisco needs to be focusing. That great beyond is going to be everyone’s target in 2020, and if Cisco wants to make something of its organizational shift, they’ll need to make sure they’re up on the firing line and ready.
How? Rival Juniper purchased HTBASE, a vendor who provided an abstraction layer to support uniform deployment across disjointed clouds and data centers, but didn’t have a clue (and still doesn’t) how to maximize the value of what they got. Cisco can’t invent an alternative to Kubernetes, or likely even promote their own flavor of or slant on it, for the same reason—that they don’t know enough. Can they buy somebody, and then let them do their thing? Not when the Kubernetes ecosystem is the most important force in Cisco’s own market. Cisco needs to bend its strategy around the hybrid cloud, which means letting the outside force win over the insiders.
That applies to their marketing and sales vision, too. Juniper and Ericsson and Nokia and Huawei aren’t really the competitors Cisco has to worry about. Traditional networking, as I’ve said in a prior blog, is moving to an open model anyway. Cisco needs to be watching VMware and Red Hat, because it’s the platform vision these vendors promote that not only provides the foundation for open-model networking, but also the foundation for profitable, higher-level, services that can reignite network spending. Cisco needs to be in on it.