Looks like there is some version incompatibility. I am getting following error.
Knime, Spark, Spark Job server each have their own separate version. What should be the latest version of each which can work together?
Table to Spark Execute Failed: An error occurred org.glassfish.jason ...
Full error scree shot attached.
Any help is highly appreicated.
ERROR Table to Spark 0:171 Execute failed: An error occured: org.glassfish.json.JsonStringImpl cannot be cast to javax.json.JsonObject
I have used docker based installation of spark-jobserver/ which is supposed to include standalone spark also
please note that we currently do not provide any support for the open-source Spark Jobserver. For usage with KNIME Spark Executor we expect everyone to use the Spark Jobserver builds provided on our website. These have been tested by us and contain some additional patches.
To download it and set it up please follow the "Installation Steps" here (there is a installation guide PDF):
Can you pl. let me know, the spark jobserver should be installed on the knime server? or should be installed in the Hadoop cluster?. our knime server is on a different box and our Clouder Hadoop cluster is on different servers.
The jobserver should be installed on an edge node of the Hadoop cluster so that it has full network access to the nodes inside the cluster and at the same time is reachable for clients (KNIME Analytics Platform, KNIME Server) connecting from outside the cluster network. For more details please take a look at the installation guide PDF on the KNIME Spark Executor website: