spark-submit.cmd

379 views
Skip to first unread message

hhx...@gmail.com

unread,
Jul 5, 2016, 10:04:30 PM7/5/16
to H2O Open Source Scalable Machine Learning - h2ostream
In windows, it looks like the spark-submit doesn't exist.

C:\spark\sparkling-water-1.6.5\sparkling-water-1.6.5\bin>C:\spark\sparkling-water-1.6.5\sparkling-water-1.6.5\bin\run-example.cmd
---------
Using example : org.apache.spark.examples.h2o.AirlinesWithWeatherDemo2
Using master (MASTER) : local-cluster[3,2,1024]
Deploy mode (DEPLOY_MODE) : client
Driver memory (DRIVER_MEMORY): 1G
H2O JVM options (H2O_SYS_OPS) :
---------
'/bin/spark-submit' is not recognized as an internal or external command,
operable program or batch file.

mat...@0xdata.com

unread,
Jul 6, 2016, 5:39:12 AM7/6/16
to H2O Open Source Scalable Machine Learning - h2ostream, hhx...@gmail.com
Have you set the SPARK_HOME env variable? What does `echo %SPARK_HOME%` print in the console?

Regards,
Mateusz

Xu HAN

unread,
Jul 6, 2016, 12:52:11 PM7/6/16
to mat...@0xdata.com, H2O Open Source Scalable Machine Learning - h2ostream
Thank you Mateusz! I'm the newbie for Spark in Windows! Do you happen to have the WINDOWS instructions for the setting up?

It only has the Unix version in the website. 



I'm trying to use the Ubuntu VMWare too. It looks like I only get this screen for Ubuntu VMWare. I'm not sure what to do next. 
lfu@ubuntu:~/sparkling-water-1.6.5$ bin/sparkling-shell 

-----
  Spark master (MASTER)     : local[*]
  Spark home   (SPARK_HOME) : /home/lfu/spark
  H2O build version         : 3.8.2.6 (turchin)
  Spark build version       : 1.6.1
----

Picked up _JAVA_OPTIONS: -Xmx10M -Xms3000M


Thank you so much for your help!

Xu

Xu HAN

unread,
Jul 6, 2016, 1:00:23 PM7/6/16
to mat...@0xdata.com, H2O Open Source Scalable Machine Learning - h2ostream
Any recommendations on the system software or set up if I'm using Windows as the primary system? I'm thinking about using cygwin. 

What system did you guys fully test on if it is a windows machine? 

Thank! Really Appreciate your help!

Xu

Tom Kraljevic

unread,
Jul 6, 2016, 2:05:40 PM7/6/16
to Xu HAN, mat...@0xdata.com, H2O Open Source Scalable Machine Learning - h2ostream

Hi Xu,

Our focus has been more on Linux and Mac, since we encounter those much more frequently in the field.
I recommend starting with the Ubuntu setup you have, that’s the well-exercised path right now.

Tom

--
You received this message because you are subscribed to the Google Groups "H2O Open Source Scalable Machine Learning - h2ostream" group.
To unsubscribe from this group and stop receiving emails from it, send an email to h2ostream+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Xu HAN

unread,
Jul 6, 2016, 2:16:00 PM7/6/16
to Tom Kraljevic, mat...@0xdata.com, H2O Open Source Scalable Machine Learning - h2ostream
Thanks! 

Which version of Ubuntu have you tested on? Did you test it inside the VMWARE or stand alone machine? I'm trying to set up a stable and well tested environment.

By the way, I'm trying to use sparkling water. Any specific system requirements?

Thanks.

-Xu


Tom Kraljevic

unread,
Jul 6, 2016, 2:39:50 PM7/6/16
to Xu HAN, mat...@0xdata.com, H2O Open Source Scalable Machine Learning - h2ostream

the internal test cluster i use most often is bare metal Ubuntu 12.04.5 LTS.

h2o-specific requirements are minimal since it’s pure java.  
recent 64-bit oracle java jdk 7 or 8 is recommended.

Xu HAN

unread,
Jul 6, 2016, 4:08:22 PM7/6/16
to Tom Kraljevic, mat...@0xdata.com, H2O Open Source Scalable Machine Learning - h2ostream
I got this in the Ubuntu 14.04 in VMWare. Do you happen to see this before?  Any idea why it has this error?

Error occurred during initialization of VM
Could not reserve enough space for object heap
Error: Could not create the Java Virtual Machine.
Error: A fatal exception has occurred. Program will exit.

Thank you! Really appreciate your patience.

-Xu

mat...@0xdata.com

unread,
Jul 7, 2016, 2:58:11 AM7/7/16
to H2O Open Source Scalable Machine Learning - h2ostream, to...@h2o.ai, mat...@0xdata.com, hhx...@gmail.com
Are you running a 32bit or 64bit Java? How are you starting H2O exactly? Setting -Xmx argument should help but whether you should set it to a lower or a higher number depends on what Java you're using. Switching to 64bit if you're using 32 would be advised anyway.

Mateusz
Reply all
Reply to author
Forward
0 new messages