Support Questions

Find answers, ask questions, and share your expertise
Announcements
Celebrating as our community reaches 100,000 members! Thank you!

Problem for adding cdsw service

avatar
Explorer

Hi,

 

I tying to enable cdsw on my cluster.

 

I followed this documentation:

https://www.cloudera.com/documentation/data-science-workbench/latest/topics/cdsw_install_parcel.html

 

I adding jar file in /opt/cloudera/csd and enable it in CM (Administration => Settings => Custom Services Descriptors )

 

I setted new parcel in host => parcels and downloaded => distributed and activated.

 

When i restart cloudera-scm-agent, i got this error:

 

 

2017-11-21 11:51:25,116 WARN main:com.cloudera.csd.components.CsdLocalRepository: Skipping [/opt/cloudera/csd/CLOUDERA_DATA_SCIENCE_WORKBENCH-1.2.1.jar]: { service.metricDefinitions[0].numeratorUnit must be present and not empty, Could no
t read service descriptor for csd [CLOUDERA_DATA_SCIENCE_WORKBENCH-1.2.1]: Could not resolve type id 'alwaysWithAny' into a subtype of [simple type, class com.cloudera.csd.descriptors.PlacementRuleDescriptor]

 

and i can't find cdsw in add service.

 

Any suggestions ?

 

1 ACCEPTED SOLUTION

avatar
Expert Contributor

The Cloudera Data Science Workbench requires Cloudera Manager 5.13+.  The error you are seeing is due to incompatibilities with earlier version of Cloudera Manager.  Please try upgrading Cloudera Manager and let us know if that doesn't resolve the issue.

 

Best,

Tristan

View solution in original post

4 REPLIES 4

avatar
Expert Contributor

The Cloudera Data Science Workbench requires Cloudera Manager 5.13+.  The error you are seeing is due to incompatibilities with earlier version of Cloudera Manager.  Please try upgrading Cloudera Manager and let us know if that doesn't resolve the issue.

 

Best,

Tristan

avatar
Explorer

Hi,

 

Thanks for your answer. Shame on me i saw that after posted my request.

 

I trying to update en 5.13.0.

 

I'll be back for give a status.

 

Regards,

avatar
Explorer

Hi,

 

I just updated CM to 5.13 and work like a charm.

 

Now another question, "my" devs use spark1.6 and i see for cdsw spark2 is needed, so is possible have 2 spark (spark 1.6 and spark2) in both on same cluster ?

 

Thanks in advance,

 

Regards,

avatar
Explorer

Hi,

 

I install spark1.6 and 2 in both on my cluster and work like a charm. 

 

I put topic in resolv.

 

Thanks all,