Rapidminer Radoop Connection with Spark 1.6.2 failed

pavan3253p
pavan3253p New Altair Community Member
edited November 5 in Community Q&A

Hi folks

 

I am trying to connect Radoop Connection with Spark 1.6.2 in standalone mode spark, Here i have configured all the required properties based on rapidminer documentation, That  configuration screenshots, I have attached below while i am performing 'quicktest', it shows the following error. But when I have selected spark version as 'None', the quicktest has succeded with warnings, but when I access to HDFS in Radoopminer HadoopData tab, it's showing the following warning. How can solve this both issues, If anyone solved please share your views

 

Thanks in Advance

 

 

 

Screenshot1 from 2017-03-03 14_04_14.pngScreenshot2 from 2017-03-03 14_04_39.pngScreenshot3 from 2017-03-03 14_04_49.pngScreenshot4 from 2017-03-03 14_05_11.pngScreenshot5 from 2017-03-03 14_12_58.png

Tagged:

Answers

  • phellinger
    phellinger New Altair Community Member

    Hi,

     

    I can see an empty entry in the Advanced Spark Parameters list. Please remove that entry, it may cause the error.

     

    Also there is a typo: mapredcue.jobhistory.done-dir instead of mapreduce.jobhistory.done-dir.

    Some of these path values are actually HDFS path. 

    But not all of these entries must be provided, most of them are retrieved from Hive automatically.

     

    Just to be sure, is YARN running on the cluster, isn't it?

     

    Best,

    Peter

Categories