CSVReader+FlowVariable+ClusterExecution

KNIME Analytics Platform
1

Hey,

I ran into a problem when executing a CSVReader node on a cluster. The Inputfile path is controlled by a FlowVariable (infile (URL)).

First it sends the warning message:

WARN      CSV Reader     Errors overwriting node settings with flow variables: Unknown variable "infile (URL)"

Afterwards the execution is terminated:

ERROR     AbstractClusterJob     Execution of node did not succeed on cluster: reason unknown

When I do not use the FlowVar to specify the Inputpath, the node executes without any errors. When try to execute it with FlowVar but without cluster execution it also works.

I also used the same FlowVar to set the Filepath of a CSVWriter, also combined with cluster execution, and it worked without any problems.

So I was wondering if this is a problem with the CSVReader or a problem on my side. Is there any suggestion to this problem ?

Thanks !

2

Dear BInfo,

I can reproduce the problem. I'll contact you by EMail to discuss this further. (The cluster execution is a commercial extension, which I guess most of the open source users are not interested in.)

Thanks,
  Bernd
 

3