I’m trying to implement error catching on my workflow. My problem is with the read excel sheet names node, the node does not fail in execution but in configuration, which does not allow the error catching node to execute.
I’ve read on the forum about using string to UI and other methods, but all these nodes require that the file is saved locally. The file that my workflow receives is an URL linked to an S3 file.
Does anyone have any ideas on how I can catch errors if the file does not exist?
I copied you an example from one of my workflows, I also use S3. What I do is, I need to check if there is a file which was created in the last 24 hours, if there is none I send an error email, if there is one I read and process it. Sorry I cannot upload this as it contains sensitive information.
Hi @Iris Thanks for your reply again, I now understand what you meant.
Unfortunately this still does not work, as I cant authenticate with Amazon. I receive S3 links from third parties to download the file. My problem is if the link to the S3 file is incorrect or expired there is now way for me to catch the error.
If you or anyone at the knime team can please let me know if you can think of another solution to the problem that would be great.
Maybe you can read the link(s) first in a table or string variable and then use a variable in the amazon node ? (can’t test it myself as I have not installed any amazon extensions)
br
if I got it right someone shares an object with you from S3 with a presigned URL which you can download without authentication? If so give GET Request node a try then. You can either continue to reading your file based on status code either configure node to fail when there is no file if you want to use Try/Catch nodes. For latter option you need latest KNIME version (4.4.0) which provides Error Handling tab in Get Request node configuration.
