HashiCorp Consul service leads off post-funding roadmap

HashiCorp drew much attention last month when it secured a $175 million round of funding on a $5 billion market valuation, cementing its place among the most prominent cloud-native software vendors. 

HashiCorp's three major open source software products -- Terraform infrastructure as code, Vault identity-based security and Consul network service discovery -- are widely used by enterprise DevOps shops as they transition from on-premises data centers to microservices workloads in the cloud. Speculation swirled about HashiCorp as an acquisition target for a bigger IT vendor for more than a year before its latest funding round, but with such a large valuation, the company seems destined for an IPO instead.

SearchITOperations.com caught up with HashiCorp CEO David McJannet to discuss how the company plans to invest the new cash investment, its outlook on Kubernetes, cloud and DevOps, and how it plans to avoid common open source business pitfalls.

HashiCorp co-founders Mitchell Hashimoto and Armon Dadgar have talked about a Nomad managed service, and there have been some SaaS offerings for Terraform. With your latest round of funding, should customers expect more things like that?

David McJannet

David McJannet: A big part of it is investing in the core [business] and go-to-market organization. We're playing an enabling role for more and more of the biggest companies on the planet, and that requires us to add more field people and architects to support that growing customer base. In terms of products, we have six products, and each of the product teams acts somewhat independently. Vagrant and Packer are early products, and not ones that we monetize, but are enormously used, and we continue to invest in them.

The other four -- Terraform, Vault, Consul and Nomad -- each have big teams, and we're looking at the adjacent markets that we can solve problems for in each space. For example, we introduced support for KMIP within Vault, which is the ability to encrypt a specific column and row in a database. That's a huge thing in the market that exists on-prem, and now it needs to be solved in cloud, and Vault is the center point for how to do that. We've also introduced the ability to do tokenization, the ability to encrypt data on the wire -- these are market adjacencies that extend the value prop of products, because over time, each of these products have become platforms. Vault is not just about solving your secrets management problem. It's actually the notion of using identity as the basis for security -- secrets management is a use case, encryption is a use case, and you can kind of spread out from there. As people go to cloud, from a vendor perspective, those categories are up for grabs.

We also plan continued investment in hosted services. We introduced Terraform Cloud in December of last year. I think of it as GitHub for ops people, and there are tens of thousands of people on that platform already. The second product that we're doing that with is Consul -- we have a managed Consul service that we announced that in the fall of last year. It'll be in public beta April 28, and it's been in private beta for months. It's an important moment for us, because what we hear over and over again is, people would like us to run those services for them. The Consul service is the first example of that. It's taken a while for us as a vendor to build a level of trust amongst the Global 2000, to the point where now they're willing for us to run that for them. That just took time.

Docker is now a kind of cautionary tale about a company that took a ton of funding and ended up not really exiting in the way that its investors expected. How does HashiCorp avoid that outcome?

McJannet: Building open source businesses is a very nuanced thing. This is my third. The other two are public companies. I was part of GitHub, Hortonworks, SpringSource, and we've learned a lot about how to build enduring open source businesses. I actually believe it's impossible to build a proprietary infrastructure company anymore. Everybody on top relies on infrastructure, and there's a desire to have the plumbing be open source, so you're not beholden to a particular vendor the way we were to Oracle a decade ago. And our business is predicated on revenue multiples -- the investors that led this our last round were public market investors, not venture capitalists who are enamored with a story. This is not an early-stage company built on the idea of having a large community. We'd like our customers in the market to understand that we have built a sizable business that the most conservative investors in the world are going to value.

I actually believe it's impossible to build a proprietary infrastructure company anymore.

Another part of that is the Cloud Engineer Certification program that we that we launched April 21. I think it's indicative of the evolving role of the cloud ops career. Years ago, people ago had this view that there's a mythical DevOps person in every company. What's become clear is that ops, security, networking and development are still the four main people in IT. We want to help those people, career-wise, that are doing ops today transition to become cloud ops engineers. There's such a massive skills gap in the market and our customers have been asking us for two years to help them find HashiCorp-certified people. This is an important part of our long-term business that's not revenue-related. If you're building long-term businesses with really strong foundations, you have to focus on building the ecosystem, and you have to focus on building out paths for the practitioner. I was at Microsoft for a long time, and their focus on the practitioner is profoundly important to building a generational franchise. That's what we're doing.

Speaking of the ecosystem, HashiCorp has a kind of 'coopetition' relationship with Kubernetes. There's some overlap between Consul and etcd, between Consul Connect and Istio, between Nomad and Kubernetes itself. How do you plan to navigate that kind of dual relationship?

McJannet: There's a lot of confusion as to where Kubernetes fits -- is it a runtime platform for my applications? Is it a new infrastructure platform like vSphere? Our view is that Kubernetes will never be the only thing that you run -- there will always be Java applications and serverless things and, you know, edge devices. We treat Kubernetes as a first-class citizen the same way we treat Java as a first-class citizen. Our view is not technology-specific -- it's focused on the compute primitives of that new cloud runtime layer. Kubernetes has actually accelerated the adoption of Terraform, Vault and Consul, because it's introduced the problems of dynamic IP addresses, perimeterless security and infrastructure on demand, which Terraform, Vault and Consul are designed to solve. But there'll be something after Kubernetes, and we'll support that too.