I am using Oracle/SQL execute snaps a lot to run some complicated queries.
Can someone explain to me:
- In my case, my query would return 100000+ documents/rows, what I found is that during the execution, the oracle/sql execution snap will first output first 2000 rows as output and gradually output abt 6 docs/s to next snap. To me it is very slow, why does it act this way? Should it output all 100000+ rows together to next snap as input?
ref pipeline: pipe_runtime_id=e791e2cb-7256-405d-af1b-92f97
- I tuned the Batch size and fetch size for my database account, but i don’t see any improvement for my query. Any guideline or best practice or example for me to follow to tune those parameters?