Big Analytics Roundup (November 2, 2015)
Spark Summit Europe, Oracle Open World and IBM Insights all met last week, as did Cloudera’s Wrangle conference for data scientists.
But in the really important news, KC beats the Mets to take the Series.
Top news from the Spark Summit is Typesafe’s announcement of Spark support, plus some insight into what’s coming in Spark 1.6. I will publish a separate roundup for the Spark Summit next week when presentations are available.
Nine stories this week:
(1) Typesafe Announces Spark Support
Typesafe, the commercial venture behind Scala and Akka, announces commercial support for Apache Spark. Planned service offerings include an offer of one day business hour response to questions for projects in development. For production, SLAs range from 4 hour turnaround during business hours up to 24/7 with one hour turnaround.
(2) More Funding for Alteryx
The New York Times reports that Alteryx has landed an $85 million “C” round, led by Iconiq Capital. That makes a total of $163 million in four rounds for the company.
(3) Oracle Adds Spark to Cloud
At Oracle Open World, Oracle announces Oracle Cloud Platform for Big Data, a PaaS offering; Dave Ramel covers the story. Key new bits include automated ingestion, preparation, repair, enrichment and governance, all built in Spark; and a DBaaS offering with Hadoop, Spark and NoSQL data services.
(4) IBM Adds Spark Support to Analytics Server
Full story here. Great news for those who want to use the high-end version of the second most popular data mining workbench with the third and fourth most popular Hadoop distributions.
(5) Ned Explains Zeppelin
Ned’s Blog provides a nice Zeppelin walk-through, noting the UI’s rich list of language interpreters, which currently includesL HiveQL, Spark, Flink, Postgres, HAWQ, Tajo, AngularJS, Cassandra, Ignite, Phoenix, Geode, Kylin and Lens.
(6) IIT and ANL Deliver BSP with ZHT
Researchers from the Illinois Institute of Technology, Argonne Labs and Hortonworks report that they have implemented a graph processing system based on Bulk Synchronous Processing on ZHT, a distributed key-value store. Nicole Hemsoth reports. The new engine, called Pregelix, when benchmarked against Giraph, GraphLab, GraphX and Hama, outshines them all.
(7) Wrangle 2015 Meets in SFO
(8) MapR Offers Free Spark Training
MapR announces availability of its first free Apache Spark course as part of its Hadoop On-Demand Training program. No word on quality, but it’s hard to beat the price.
(9) Cloudera Pushes HUE for Spark
On the Cloudera Engineering blog, Justin Kestelyn explains how to use HUE’s notebook app with SQL and Spark.