Hortonworks will become a subsidiary of Cloudera and will fill four seats on the nine-member Cloudera board. Cloudera CEO Tom Reilly will stay in charge.
The merger comes as cloudy future faces Hadoop technology, which is now squeezed by the growing popularity of cloud computing. Hadoop relies on batch processing and risks becoming a dinosaur in the age of real-time data manipulation. Business intelligence ages quickly and Hadoop’s slow response cycle reduces its utility.
Reilly has said that the combined companies will now form an enterprise data hybrid cloud platform even as legacy Hortonworks and Cloudera platforms will continue to be supported for the next three years.
Alex Hickey reported on the merger for CIODive:
The Hadoop premise was taking data from everywhere and making big data, but this left many businesses with piles of data they weren’t doing anything useful with, said Yaron Haviv, founder and CTO of Iguazio, in an interview with CIO Dive. Now, data is leveraged across business applications in the short term to generate business value. The value of data diminishes over time, so businesses need to be able to act on it quickly. This is where cloud capabilities and advanced computing techniques come in.
But a less rosy view of the merger would take into account the declining prevalence of the Hadoop framework, an open-source software for data storage, orchestration and computing. It’s a 10-year-old technology, and many of its layers are being cannibalized by the cloud, Haviv said.
At the storage layer, more customers are moving their data to the cloud; offerings such as Amazon Web Services S3 are cheaper, more scalable and fully provisioned, he said. And at the orchestration layer, frameworks such as Kubernetes are more general purpose than Hadoop. Finally, at the computation layer, many cloud providers are also providing functions such as AI and ML to help customers manipulate their data.
