Packing some trunks

SAP to sell Hortonworks and Intel-flavoured Hadoop

Elliot Bentley

German data giant SAP certified third-party distros for in-memory database tech HANA, with more to come before end of year.

German data giant SAP is to redistribute other vendors’ Hadoop distributions, the company announced today.

A partnership with Hortonworks and Intel will see two flavours of the big data platform certified for and redistributed with SAP’s HANA in-memory database tech, allowing for – if the press release is to be believed – “real-time insights with extreme storage”.

It follows the announcement in February from SAP and Intel that the two had formed a then-undefined “technical and business partnership”, integrating Intel’s Hadoop tech with HANA.

Today at TechCrunch Disrupt, the fruits of this coupling were revealed. Rather than rolling its own version of Hadoop, SAP will offer customers a choice between two very different distributions, Intel’s performance-optimised proprietary interpretation or Hortonworks’ close-to-the-trunk open source version.

The company says it plans to certify additional distributions in the future, with more to come by the end of the year – a clever move, especially in a market still without a clear leader.

As larger companies have begun lumbering into the rapidly-expanding Hadoop market, they’ve often opted to partner with existing, smaller players more familiar with the technology. Both Google and Amazon chose to integrate MapR’s Hadoop distributions into their cloud platforms, and business intelligence supplier Jaspersoft enticed both Cloudera and Hortonworks into joining their “Hadoop Partner ecosystem”.

Elsewhere in the industry, Hadoop’s open-source roots have been embraced by many of its most active players. Hortonworks and Cloudera open source much of their research, such as Impala or a faster Hive, Cloudera and Twitter have partnered on Parquet, a new columnar storage format for HDFS, and even Microsoft has got in on the act with its REEF Framework.

Inline Feedbacks
View all comments