Hortonworks introduces DataFlow, acquires Apache NiFi-backer Onyara
Hortonworks is betting big on Apache NiFi, acquiring Onyara, the commercial entity behind it, and releasing DataFlow, a separate subscription alongside the Hadoop-based Hortonworks Data Platform.
Context & Ripple Effects
This is the second bolt-on Hortonworks has made in 2015, and it follows a recognizable template: weeks after buying SequenceIQ for rapid deployment tooling and pledging to donate the tech to Apache (acquired SequenceIQ), the company is again absorbing an open-source project's commercial entity rather than building in-house. The prize this time is Apache NiFi via Onyara, repackaged as Hortonworks DataFlow and sold as a subscription sitting alongside the Hortonworks Data Platform.
The strategic logic is visible against the rest of the corpus: the Hadoop storage-and-processing layer was being surrounded by adjacent data-movement plays, with Confluent raising a $24M Series B weeks earlier around its Kafka-based stream platform ($24M Series B) and Pivotal, IBM and others forming the Hadoop-focused Open Data Platform in February. Hortonworks needed a flow-and-ingestion story so its platform could claim more than batch processing.
First-order effects
- Hortonworks now sells two subscriptions instead of one — DataFlow beside the Hortonworks Data Platform — giving it revenue tied to data movement and routing rather than Hadoop compute alone.
- Onyara's team and the NiFi roadmap move under Hortonworks' stewardship, and given the SequenceIQ precedent, some portion of the technology is likely headed toward Apache Software Foundation governance.
Second-order effects
- Confluent and other stream-platform vendors gain a direct competitor backed by an established Hadoop distribution, forcing the data-pipeline market to compete on integration with Hadoop estates rather than standalone adoption.
- Rival Hadoop distributors must decide whether to match with their own flow/orchestration acquisitions or cede the ingestion-and-routing attach sale to Hortonworks' sales motion.
Third-order effects
- If the pattern holds, Hadoop vendors stop being storage-and-compute companies and become multi-product data platforms assembled through repeated open-source acquisitions — a trajectory the corpus shows culminating three years later when Hortonworks and Cloudera announce their $5.2B all-stock merger (announced all-stock merger).
- It also foreshadows the automation of the pipeline work itself: Ascend's later funding round for an autonomous dataflow service that reduces data-engineer effort (autonomous dataflow service) points to flow management becoming a product category independent of any single Hadoop vendor.
The trend: Hadoop-era infrastructure vendors are capturing the workflow and data-flow layers above the platform through open-source acquisitions, converting commodity storage-and-compute into higher-margin orchestration subscriptions.