Closed exalate-issue-sync[bot] closed 1 year ago
Jakub Hava commented: Closing as this is really old. I suggest trying the latest versions where a lot of improvements are available and reporting a new jira if necessary
Jakub Hava commented: Closing as this is really old. I suggest trying the latest versions where a lot of improvements are available and reporting a new jira if necessary
JIRA Issue Migration Info
Jira Issue: SW-114 Assignee: UNASSIGNED Reporter: Michal Malohlava State: Closed Fix Version: N/A Attachments: N/A Development PRs: N/A
JIRA Issue Migration Info Cont'd
Jira Issue Created Date: 2016-04-08T10:55:55.986-0700
I've never had a problem using the old documentation (below) starting up a Spark EC2 cluster using the Spark EC2 script, then SSH'ing in, downloading Sparkling Water, then launching an H2O cloud on the Spark EC2 cluster, but I can't get it to work under the "new" documentation on the website.
Once I have a Spark 1.6.1 cluster running (let's say the master is 12.34.56.78) what is the proper script to launch an H2O cloud? When I try the new script (per the website) or my old one (below) I get a variation of this error (my emphasis):
So assuming I start up a Spark cluster using the script below, then SSH in, how do I get an H2O cloud running on the Spark cluster?
Old Code Begin
wget http://h2o-release.s3.amazonaws.com/sparkling-water/rel-1.4/10/sparkling-water-1.4.10.zip
unzip sparkling-water-1.4.10.zip
rm sparkling-water-1.4.10.zip
Start Sparkling Water
cd ./sparkling-water-1.4.10/
export SPARK_HOME='/root/spark/'
export MASTER='spark://ec2-12-34-56-78.compute-1.amazonaws.com:7077'
bin/sparkling-shell
import org.apache.spark.h2o. val h2oContext = new H2OContext(sc).start(1) import h2oContext.
Old Code End
--