IBM Buying HashiCorp: What Devs, Analysts and Competitors Are Saying
As one might expect, sentiment in the HashiCorp and OpenTofu communities about the future of the popular Infrastructure as Code (IaC) software tool Terraform after the company’s impending acquisition by IBM is mixed.
Big Blue, which announced April 24 that it will pay $6.4 billion ($35 per share) for HashiCorp, has stated that the company will remain independent after the takeover, but based on what actually happened after its previous acquisitions in the IT world, not everybody is sold on that promise.
While some members of the community have expressed frustration and disappointment with HashiCorp’s decision in advance of the sale to adopt the proprietary Business Source License (BSL) in place of the standard Mozilla Public License v2.0 (as of August 2023), others saw it as an opportunity for the community to take control and create their own fork of the project.
For the record, OpenTofu is a forked open source Infrastructure as Code tool developed as a community-driven alternative to Terraform. It was created in response to HashiCorp’s decision to adopt the BSL, which changed Terraform from open source to proprietary source available.
Now hosted by the Linux Foundation, OpenTofu allows users to manage cloud and on-premises resources through human-readable configuration files. It supports a wide array of services via a public registry and functions with a plan-apply cycle, using a state file to track resource states.
What Developers Are Saying
Naturally, there is an overwhelming desire among community members to maintain Terraform as an open source project. But OpenTofu is rapidly gaining popularity, with over 31,000 downloads registered in December 2023. A segment of the community has published a manifesto calling on HashiCorp to reverse recent changes to the licenses governing the usage of Terraform. A group within the community is actively advocating for the project to remain open source.
Some community members have expressed concerns about restrictions that HashiCorp has put in place ahead of the takeover, such as limitations on how Terraform can be used in production environments. Others have questioned the move overall and its potential impact on the future of the platform, which has great value as a cross-cloud development platform.
Overall, the sentiment in the HashiCorp community about moving Terraform back to open source is a mix of frustration, disappointment and a renewed desire for open innovation and competition.
Questions about the IBM acquisition plans are not being discussed publicly in Terraform community forums at this time. This is not uncommon during such acquisitions.
Developer Robert Hafter said on one of the OpenTofu forums that “Oracle has officially announced it, they’re running OpenTofu internally and have switched their Cloud Manager service to using it as well. I know there are other companies who have been less vocal about it but have still moved.” So the OpenTofu movement appears to be gaining serious momentum.
More Perspectives from Developers
Here are more perspectives from developers who use Terraform and/or OpenTofu:
“Terraform has been the de facto standard in IaC for years, but HashiCorp investment in the codebase was incredibly disappointing,” developer Troy Knapp wrote in a OpenTofu forum. “Important issues were left open, and valid PRs went unreviewed for years. It became increasingly apparent that their focus was on their cloud offering and not on the language itself. Once Terraform was forked the number of full-time core maintainers in OpenTofu was 5X the staff in TF. Issues that were long neglected started immediately being addressed and the roadmap was no longer a black box, but dictated by community involvement. My primary worry is that if Terraform goes back to open source, then investment in the language will go back to status quo. My hope is that IBM works to merge OpenTofu and Terraform in the Cloud Native Computing Foundation and solicits further investment from AWS, Azure and Google to at least match the current investment that companies like Spacelift, Scalr, and Env0 are making in OpenTofu.”
“If IBM will buy HashiCorp, the expectation is that the license will change. But different people expect (or are afraid of) different charges. There are two options — revert back as it was or go GPL (GNU Public License). The former feels like a status quo, the latter is unlikely, and it would create some interesting situations. Since Terraform as a program is standalone, it wouldn’t affect much on the first sight because using GPL software doesn’t ‘infect’ you with GPL, linking and code reuse does. But that would mean that OpenTofu and Terraform would stay separated due to license incompatibility. On myself: I’m developing a proprietary Terraform provider (plugin). I work for a large public U.S. company,” — wrote developer Konstantin in an OpenTofu forum.
“Right at the beginning of the license change, I knew HashiCorp is going to sell itself, if you consider the disputed lawsuit from MongoDB, the license change is from the suggestion of MongoDB investor or quite related, but they wouldn’t replicate the same success as Mongo, ’cause the community won’t be fooled twice from my open source experiences. In short, there is no room for HashiCorp to play now, OpenTofu will go ahead and win, Terraform has no foundation to support its development speed as before,” wrote developer Hao Wang in an OpenTofu forum.
“My company has to decide whether or not we are migrating to OpenTofu. Some people were hoping HashiCorp would revert their decision after the success OpenTofu is having. As a practitioner, we have been holding back on the adoption of new Terraform versions to ensure we are capable of migrating when management makes a call on OpenTofu. For personal projects, I have fully migrated. The decision to migrate to OpenTofu won’t be a companywide decision but more of a team/business unit decision, in particular since my team doesn’t leverage any enterprise features,” developer David Gamba posited in an OpenTofu forum.
What the Analysts Are Saying
“IBM’s acquisition of HashiCorp may raise concerns among developers, particularly within the open source community, regarding the impartiality and independence of HashiCorp’s services,” says Paul Nashawaty, practice lead at The Futurum Group, “While the move aligns with IBM’s focus on multicloud solutions and enhances credibility in this area, there’s a risk of perceived bias undermining HashiCorp’s appeal. However, the acquisition also presents opportunities for IBM to leverage HashiCorp’s multicloud capabilities to strengthen its position in the market and provide value to enterprise clients.”
Nashawaty said the impact of this acquisition will need to be well-managed across the IBM, Red Hat portfolio (IBM acquired Red Hat in 2019).
On the one hand, if IBM keeps HashiCorp separate and open source, I can see the collaboration between the portfolio being complementary and allowing customers to choose the tech stack appropriate for their needs, similar to what they need today. On the other hand, forcing or merging the tech stack together will limit customers’ choices,” Nashawaty told The New Stack.
Constellation Research Analyst Dion Hinchcliffe told The New Stack that “HashiCorp is the latest in a long spree of acquisitions over the last year by IBM CEO Arvind Krishna to round out their cloud offerings to make them more competitive with the hyperscalers. HashiCorp has struggled at times to crack the enterprise sales market, despite being one of the cooler companies on the block. While IBM will undoubtedly use HashiCorp’s strong developer ‘street cred’ as a proof point in its own offerings, it remains to be seen if HashiCorp can retain its perceived neutrality as a cloud infrastructure software provider at a time that robust multicloud and cross cloud support continues to grow in importance.”
Fellow Constellation analyst Holger Mueller had a more cut-and-dried take: “HashiCorp does not make sense for IBM. The services model lives from being independent, and now they may look biased. And the service revenue around DevOps is going to dry up. The multicloud aspect of HashiCorp makes sense from an IBM credibility perspective,” he told The New Stack.
Perspective of a Competitor
Sachin Aggarwal, CEO and co-founder of appCD, a developer of generative infrastructure from code, had the following opinion on what this means for Infrastructure as Code (IaC) integration: “Infrastructure as Code (IaC) has played a crucial role in driving the widespread adoption and expansion of cloud technologies. However, despite its significance, innovation within the realm of IaC has seen slower progress, a trend that is likely to persist until the implications of HashiCorp’s integration into IBM become more clear. Until then, the persistence of manual processes and the potential for errors in IaC implementation will continue to impose significant cognitive burdens on engineering teams.
“HashiCorp’s historical challenges in delivering a truly exceptional developer experience and robust governance at scale pose significant considerations for IBM as it seeks to monetize and expand HashiCorp’s offerings within its customer base.”
More on this to come.
