Post
by peeterskris » Fri Dec 06, 2013 1:39 pm
Did they mention anything about their "SQL-on-Hadoop" offering? I'm a bit skeptical if it's really SQL-On-Hadoop. Do they store their data in HDFS? Do they use Hadoop nodes to do the processing? Of course not MapReduce. But, like Impala or Presto, still really run on Hadoop? Or do they have a connection with Hadoop and do everything on their own nodes?
That's a big difference in my opinion. The first option will slow down Vertica because the way HDFS is built. The second option is not really SQL-On-Hadoop.