I am using using GenerateTableFetch
with incoming flow files containing table_name
to import data. And I have put Additional WHERE clause
to handle incremental updates using updated_at
column of the table manually.
Issue is when there is no data to import for the constructed query, it simply drops the flow file. It should have a separate relationship other than success and failure for this case so that I can handle on my own if something has to be done for such cases.
Currently, at the start of the workflow, I am using Wait
on number of tables I have put into flow and tracking when they are done to increment the done count using Notify
. But for the tables which get's dropped in GenerateTableFetch
, it just keeps on waiting.
Any way to handle this?
This would indeed be a good improvement to add to GenerateTableFetch, perhaps an optional property to "Output Empty FlowFile on Zero Results" or something. I have written NIFI-5604 to cover this improvement. Combining this with NIFI-5601, you could do routing downstream (if fragment.count = 0 for example).