LAE

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: aop

Hi!

Is it possible to create scenario similar to using the bypass node but with the difference that the input availability decision would be based on node run status (processed succesfully/error)rather than its enabled/disabled status? In other words I would like the output to be input1 if all nodes further down the stream processed succesfully and input2 if there was an error in some node further down the stream of input1.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: stonysmith

I assume that by "downstream" you mean "predecessor nodes".. usually referred to as "upstream".

Unfortunately, there's no such thing as a "soft error" at this time.
If a node results in an error, then any successor nodes (downstream) are not executed.

===
What you COULD do is break the processing up into three separate graphs: Data1, Data2 and Analysis. When Data1 and/or Data2 execute successfully, they write out their main data files. Then, in the Analysis graph, use the Directory node to look for those two files an use a filter to decide which file to process.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: aop

You are correct, that's what I meant.

I was just looking for a way to e.g. be able to create an excel output with multiple sheets succesfully every time even if some of the data streams going to excel output run into node error. For the error cases I would have liked to bring an error message via input2 in bypass or something like that. This way the excel output could always be created and an error message would be present on each sheet that ran into a node error during the analysis.

So to be exact I think my problem in this scenario is not the lack of "soft error" (as my second input would be a completely different stream altogether) but the fact that the bypass node considers input to be 'not available' only when some predecessor node is disabled even though in my scenario I could say that in case of node error in predecessor nodes that specific input could also be considered as not available. However, I understand the current behaviour of the bypass node is by design. I was just hoping there would have been some workaround.