In Q1 2026, a five-person DevOps team opened a renewal email from Terraform Cloud. Their annual bill had jumped from $180,000 to $270,000. A 50% increase with zero new features to show for it. The team lead forwarded the email to Slack with one line: “We need to talk.”
Three months later, that team had migrated 600 Terraform modules to OpenTofu, replaced Fivetran with self-hosted Airbyte for their data pipelines, and swapped Datadog for Grafana plus SigNoz. Annual infrastructure tooling spend dropped 58%.
This is not an isolated story. Across North American B2B SaaS, a structural shift is underway: open source infrastructure is systematically displacing commercial platforms that once seemed permanent. The catalyst isn’t ideology. It’s economics, control, and a growing sense that vendor loyalty is a liability.
The Trust Collapse
The timeline matters.
In August 2023, HashiCorp changed Terraform’s license from Mozilla Public License (MPL 2.0) to Business Source License (BSL 1.1). BSL is not an OSI-approved open source license. It prohibits anyone from using Terraform to build a “competing product,” and the definition of “competing” is vague enough to keep legal departments up at night.
The community responded fast. By September 2023, the Linux Foundation announced OpenTofu, a fork based on the last MPL-licensed version of Terraform. IBM, Gruntwork, Spacelift, and others immediately backed it.
Then in late 2024, IBM acquired HashiCorp for $6.4 billion. Terraform, Vault, Consul, Nomad: all became IBM assets. IBM needs to recoup that investment. Tighter enterprise licensing, feature gating, and ecosystem lock-in are the predictable commercial path forward.
This pattern is not new. Elastic switched Elasticsearch from Apache 2.0 to SSPL in 2021; AWS forked OpenSearch within weeks. MongoDB did the same thing in 2018. Elastic eventually added an AGPL option in 2024 under community pressure, but trust, once broken, costs more to rebuild than it ever cost to maintain.
As of April 2026, 38% of Terraform users report they are evaluating or actively migrating to alternatives. OpenTofu’s annual downloads grew over 300%, with cumulative downloads approaching 10 million. GitHub stars reached 29,300 by June 2026.
The logic is straightforward: when your infrastructure tooling can change terms unilaterally, you either accept that risk or you take back control.
The Cost Math
Cost is the primary driver of migration. More than philosophy, more than principle, the spreadsheet wins.
Consider data integration. Fivetran charges by Monthly Active Rows (MAR). A mid-size SaaS company processing 50 million rows per month typically pays $120,000 to $180,000 annually. Switching to self-hosted Airbyte brings infrastructure costs (Kubernetes cluster plus storage) to roughly $20,000 to $40,000 per year, plus about half an engineer’s time for maintenance. That’s a 50-70% reduction.
One e-commerce platform saw its Fivetran bill triple in a single quarter because of a single high-frequency table updating 8 million rows daily. After migrating to Airbyte, the same data volume produced only infrastructure costs. No per-row billing surprises.
Observability shows an even wider gap. Datadog prices on a per-host, per-module SKU model. Infrastructure monitoring runs $15-23 per host per month, APM adds $31 per host, log management starts at $0.10 per GB, and Synthetics charges per test run. A 50-host team running Infrastructure plus APM plus Logs easily exceeds $5,000 monthly. Over $60,000 annually.
Self-hosted SigNoz, built on ClickHouse and OpenTelemetry, covers logs, traces, and metrics for that same 50-host footprint at $800 to $1,500 per month in cloud compute and storage. Even accounting for engineering maintenance time, total cost runs 20-30% of the Datadog equivalent.
For IaC specifically, Terraform Cloud’s Team and Governance tier starts at $70 per user per month. Enterprise pricing generally exceeds $100 per user. OpenTofu paired with open source CI/CD pipelines (GitHub Actions, GitLab CI) brings the tooling layer cost to effectively zero.
Production Readiness in 2026
Two years ago, choosing open source alternatives required a leap of faith. Connector gaps, thin documentation, small communities. Those concerns were legitimate.
The landscape has changed.
OpenTofu has reached version 1.12.x and diverged meaningfully from Terraform’s codebase. Features unique to OpenTofu include native state file encryption (supporting AWS KMS, PBKDF2, and other key providers), dynamic provider configuration, loopable import operations, and provider iteration. Several of these capabilities require paid Enterprise versions in the Terraform world, if they exist at all.
Migration remains smooth for now. Both projects share compatible core language syntax and state file formats. Existing Terraform modules run on OpenTofu with minimal or zero modifications. But this window is narrowing. As each project introduces exclusive features, future migration complexity will increase.
Airbyte now offers over 350 connectors in its open source version, covering major SaaS platforms, databases, and API sources. The Connector Development Kit lets Python developers build custom connectors in one to two days. Airbyte Cloud provides a managed option at pricing typically 30-50% below Fivetran.
For comparison: Fivetran, which completed its merger with dbt Labs in June 2026, maintains over 500 connectors with higher consistency (all maintained by paid engineers). But Airbyte’s advantage lies in flexibility. Self-hosting means data never leaves your network, which is a hard requirement for regulated industries.
Grafana plus SigNoz covers the three pillars of observability: metrics, logs, and distributed traces. SigNoz natively supports OpenTelemetry and uses ClickHouse for storage (the same columnar database running at Uber’s scale), delivering strong query performance on large datasets. Grafana’s dashboard ecosystem and alerting capabilities are already industry standard. Together, they cover Datadog’s core use cases.
The Composable Stack vs. the All-in-One Platform
Fivetran and dbt Labs completed their merger in June 2026, now serving over 100,000 data teams. The rationale: end-to-end data infrastructure. Data movement plus transformation plus governance, one platform.
This is precisely the model the open source camp rejects.
The composable architecture philosophy says: pick the best tool for each layer and use open standards as the connective tissue. SQL for queries. OpenTelemetry for instrumentation. Apache Iceberg for storage format. Airbyte for ingestion, dbt Core (still open source under Apache 2.0) for transformation, Dagster or Prefect for orchestration, DuckDB or ClickHouse for analytics.
The benefits are concrete. First, no single point of failure. A Fivetran outage takes down your entire pipeline; self-hosted Airbyte’s availability depends on your own Kubernetes cluster. Second, pricing transparency. You pay for compute and storage, not for “platform.” No MAR multiplier producing surprise invoices. Third, swap flexibility. When a better tool emerges, replacing one component costs far less than replacing an entire platform. OpenTelemetry’s OTLP protocol means your instrumentation code works unchanged across backends. Switch from SigNoz to Grafana Tempo, or the reverse, without touching application code.
The tradeoff is real: you own the integration layer, version compatibility testing, and operations. This is not free. It is a trade of engineering time for control.
Why Open Source Moves Faster
OpenTofu delivered multiple features that Terraform left unshipped for years. Client-side state encryption was an RFC submitted to Terraform in 2016. OpenTofu landed it in under a year after the fork.
The difference is governance. OpenTofu operates under the Linux Foundation with a Technical Steering Committee that meets biweekly. Community members participate directly in RFC discussions and code contributions. Terraform, under IBM, follows a roadmap driven by commercial priorities. New capabilities are more likely to appear in paid Terraform Cloud or Enterprise than in the open source edition.
Airbyte’s connector development model shows the same dynamic. The CDK lowers the contribution barrier so community developers can respond quickly when a SaaS product updates its API. When Notion ships a new API version, community patches often land faster than the next scheduled release from a commercial vendor.
Grafana Labs provides the textbook example. The company keeps core products (Grafana, Loki, Tempo, Mimir) open source while generating revenue through Grafana Cloud and enterprise support. This model proves that open source and commercial sustainability can coexist, as long as you don’t treat the community as unpaid beta testers.
Where Open Source Falls Short
The open source narrative frequently glosses over several realities that enterprise buyers know firsthand.
Operations cost is not zero. Self-hosted Airbyte requires Kubernetes cluster maintenance: upgrades, scaling, incident response. OpenTofu state management needs a self-built remote backend. SigNoz’s ClickHouse cluster requires tuning as data volumes grow. These tasks need dedicated staff. A 20-person startup without a platform engineering team may find that Airbyte’s hidden operational costs exceed what Fivetran would have charged.
SLAs and commercial support carry genuine value. Datadog guarantees 99.9% uptime with a dedicated technical account manager. Self-hosted SigNoz uptime is your responsibility. At 3 AM when ClickHouse runs out of memory, nobody is on call for you. For financial services, healthcare, and other regulated sectors, a vendor’s SOC 2 Type II certification, HIPAA compliance, or FedRAMP authorization are not optional line items.
User experience gaps exist. Fivetran’s automatic schema drift handling, anomaly detection, and retry mechanisms work out of the box. Airbyte is less polished on these details. Datadog’s Session Replay, Synthetic Monitoring, and Real User Monitoring form a digital experience suite with no direct open source equivalent.
Not every organization can self-host. A data team with three analysts and zero infrastructure engineers cannot realistically operate Airbyte on Kubernetes. Fivetran’s “zero maintenance” value proposition is real for these teams. They are paying for the insurance of not having to worry.
A Practical Migration Playbook
The smart strategy is not “all open source” or “all commercial.” It is understanding where your organization’s capability boundary sits.
Start with three questions. How much engineering bandwidth can you dedicate to infrastructure maintenance? If the answer is “almost none,” fully self-hosted may not fit. Airbyte Cloud or SigNoz Cloud offer a middle path: open source core, managed operations.
How strict are your data compliance requirements? If data cannot leave specific geographic regions or network boundaries, self-hosting becomes nearly the only option regardless of cost.
What does your growth curve look like? If data volumes triple annually, usage-based commercial SaaS gets progressively more expensive while self-hosted open source scales at near-linear marginal cost (primarily storage and compute).
Phase your migration rather than cutting over all at once. Validate in non-critical environments first. Manage dev infrastructure with OpenTofu. Sync a few non-core data sources through Airbyte. Monitor a handful of microservices with SigNoz. Build operational confidence before expanding to production.
For Terraform-to-OpenTofu specifically, the migration window is still favorable. Core language and state file compatibility keep risk low and engineering effort small. But as both projects diverge further, the cost of waiting only increases.
What 2027 Looks Like
Open source infrastructure adoption will accelerate over the next 18 months. Three forces drive this. First, AI workloads are creating intense cost pressure. GPU spending already dominates budgets; observability and data pipelines should not consume what remains. Second, multi-cloud is becoming default architecture, which amplifies vendor lock-in risk. Third, platform engineering as a discipline is maturing, meaning more companies have internal teams capable of operating their own tool stacks.
The Fivetran and dbt Labs merger is a defensive move by the commercial camp: bundling to reduce switching motivation. IBM will almost certainly gate more Terraform features behind paid tiers, which will push OpenTofu adoption further. Datadog’s 15-plus product SKU model is being systematically unbundled by the OpenTelemetry ecosystem.
Open source will not fully replace commercial SaaS. Linux did not kill Windows. But it is resetting the baseline for what “reasonable pricing” and “acceptable control” look like. When an open source stack covers 80% of the functionality at 20-30% of the cost, the remaining 20% feature gap requires a very compelling justification from commercial vendors.
Ultimately, this is not a technology decision. It is an organizational capability question. Can you operate the stack you choose? If yes, open source gives you freedom and economics. If no, commercial SaaS sells you the service of not having to think about it. Both are valid answers. The important thing is being honest about which side of that line you stand on.
Stay updated with our latest AI insights
