The CSV is about 14 Mb ( about 180K very wide records ). The CALL alone is 4 Minutes. The UNLOAD alone is 10 Seconds. However, if I run the UNLOAD query immediately after the CALL, it takes 2 Minutes. The CSV is the same size each time. Very odd

What kind of hardware are you running on?

Is the cache big enough?

What exactly does the table look like?

What exactly does the UNLOAD look like?

14M and 800 bytes per record should not take even 10 seconds, on a modern consumer-grade desktop.

A test of a million row table unloaded to a 65M CSV file ran in less than two seconds.

CREATE TABLE DBA.inventory ( -- 1,000,000 rows, 30M total = 27M table + 2.1M ext + 356k index, 32 bytes per row
   item_id         /* PK        */ INTEGER NOT NULL DEFAULT autoincrement,
   item_count                      INTEGER NOT NULL,
   item_name                       VARCHAR ( 1024 ) NOT NULL,
   updated_at                      TIMESTAMP NOT NULL DEFAULT timestamp,
   last_modified                   TIMESTAMP NOT NULL DEFAULT timestamp,
   CONSTRAINT ASA78 PRIMARY KEY ( -- 356k
      item_id )
 );

UNLOAD TABLE inventory TO 'c:\temp\inventory.csv' FORMAT TEXT;
Execution time: 1.895 seconds

 Directory of C:\Temp
05/29/2017  02:12 PM        66,566,964 inventory.csv

I see it now. It is not the UNLOAD. Any query I run immediately after the Stored Procedure takes 2 Minutes, even "SELECT 1". The Procedure runs three different Cursors, looping thru 186K Records, that set multiple Fields. Perhaps I am missing a COMMIT somewhere, and the Procedure needs 2 Minutes to clean up ?