Original URL: http://www.theregister.co.uk/2011/12/02/couchdb_/
NoSQL hopeful cozies up to Hadoop data-muncher
Big data love-in
NoSQL data store CouchDB has become Hadoop’s latest convert with delivery of a connector tying together the two big-data architectures.
CouchDB user Couchbase has announced a certified Couchbase Hadoop Connector, developed with Hadoop shop Cloudera.
The connector potentially simplifies movement of data between the Couchbase Server, which Couchbase says is "powered" by CouchDB, and the Cloudera Distribution including Hadoop (CDH). Couchbase uses capabilities of CouchDB such as mobile and sync. Both CouchDB and Hadoop, meanwhile, are Apache Software Foundation (ASF) projects.
The connector does this using Sqoop, a plug-in that's an Apache incubator project. Sqoop can stream data from the Couchbase system to the Cloudera Hadoop distribution.
According to Couchbase, the Sqoop addition will enable both consistent application performance and heavy MapReduce data-crunching of data sets.
The plug-in targets web applications such as ads targeting, where low-latency is needed along with high throughput.
The Hadoop Connector was certified via Cloudera’s Certified Technology program.
Couchbase wasn't the only one cozying up to Hadoop this week. Open-source enterprise data integration start-up Talend announced that version 5 of its suite features enhanced support for Hadoop's data warehouse Hive, the Pig data analysis tool and Sqoop.
Couchbase is closely aligned with the NoSQL crowd. It uses and supports the CouchDB document store with employees such as Jan Lenhardt still contributing to the CouchDB Apache project. Couchbase also utilizes Memcached in its Membase Server. The company's customers include the BBC and Zynga.
CouchDB recently made the news when it was dropped from Canonical's Ubuntu One service after Canonical tried and failed to make the document store scale to millions of users and databases over a period of three years.
Cloudera is home to Hadoop founder Doug Cutting. Hadoop was inspired by Google’s MapReduce for large-scale data-munching.
Cloudera was first to deliver productisation and support for Hadoop, but was this year joined by Yahoo! spin out Hortonworks. The latter has this year worked with Microsoft to develop a Hadoop plug-in for Microsoft’s SQL Server, a move that has ended the life of Microsoft’s own big data cruncher, Dryad. ®
This article has been updated to clarify Couchbase's use and support for CouchDB.