Skip to:

  1. Skip to Jira Navigation
  2. Skip to Side Navigation
  3. Skip to Main Content

Cannot use Null for UDT

Description

It appears that saving POJOs with NULL UDTs is disallowed when they shouldn't be. Relevant stack trace:

http://pastebin.com/tdGeSkxM

Environment

None

Pull Requests

https://github.com/datastax/spark-cassandra-connector/pull/1153 - b1.6 https://github.com/datastax/spark-cassandra-connector/pull/1154 -b2.0

relates to

Activity

Show:

Stefan Miklosovic December 4, 2017 at 7:51 PM
Edited

this was not fixed fully, UDTs are covered but some stuff is still missing. I have created separate JIRA for it, my way of fixing it is here

https://datastax-oss.atlassian.net/browse/SPARKC-517

Jaroslaw Grabowski November 29, 2017 at 12:34 PM

+1

Russell Spitzer November 29, 2017 at 5:00 AM

A reproduction was given in the Datastax Academy chat and I've written a patch.

Russell Spitzer January 25, 2017 at 8:38 PM

Please update the environment. There is no information of the version of C* or The Spark Cassandra connector. Currently this seems to be ok in b1.6 at least for the generic CassandraRows. So please paste all of the relevant version information and if possible a sample of code producing the error.

Anonymous September 26, 2016 at 9:44 PM

FWIW - using an instantiated object for the UDT but leaving every field inside of it unset works pretty well. I wasn't able to figure out a clean solution to the original exception, since it appears that the binding between Object type -> Converter type inside of the connector is happening much earlier in the process and is agnostic of the value of the field itself.

Fixed

Details

Assignee

Fix versions

Reviewer

Components

Affects versions

Priority

Created September 22, 2016 at 12:17 AM
Updated February 13, 2018 at 3:56 PM
Resolved November 29, 2017 at 8:19 PM

Flag notifications