How long is this workflow running?
Could it be that it is running for a long time and the connection to the SMB file share is lost?
Within the Node we do have timeout parameter where you can try to set to a higher value and see if it helps to resolve the issue.
Is the issue is occurring intermittently or every time when the wf is running?
Also it would be good to rule out disc space and permission issues (check if there is enough space to write the data and whether you have the permission to write to that folder.)
If none of the mentioned above helped:
Disk share connection is linked with the session, seems like the session to the share is getting closed before the write operation. You might need to introduce some kind of timeout or refresh your session before it writes to Disk share. Is it possible to check if there is an option on your SMB drive(shared drive) to refresh the session?
I am running into the same Error. (Execute failed: DiskShare has already been closed). Looks like “session gets closed” is the problem here. How can I deal with it?
The timeout value is definetley not responsible for that. If I do an operation and then wait more than the timeout the SMB connection is still there. But if I wait much longer (e.g. 30 min) then the DiskShare gets closed.
This is a big issue for my workflow because it looks like this:
SMB Connector → List Files → Loop over each file and process it
The problem is that the processing reads the file (this is fast) and then takes, depending on the file, something between 10 and 80 minutes to process it. So the workflow does a few loops until it encounters a file which takes longer to process and then fails. (After reading the file, there is no more access to SMB until reading of the next file).
So how can I “refresh” the session? The DB Connectors have an option to automatically reconnect, but the SMB node does not.
I was thinking of doing a separate loop like “wait 1 min” → access a file → loop again. This would probably keep the session alive. But How would I end this loop when my processing is done? (also, this is ugly as hell).
Any better ideas?
PS: the SMB server is probably not the problem here. If I let Windows do the SMB connection and use a local filesystem connector, everything works fine.
is this happening when you run this workflow on the KNIME Server?
Is the workflow running locally on your AP?
Or do you experience the same (Execute failed: DiskShare has already been closed) running the workflow locally and the KNIME server as well?
Also it is possible to share the SMB Connector settings that you use in the workflow (screenshots of all tabs)?