Re: Table I/O Doesn\'t Work
Ok, I am looking this topic over for the - errr about 12th time, trying to see if there is anything I can contribute.
Andy, first you need to be consistent about the problem. You talk about insert/fetch then about F0911.update, then about F4311.FetchSingle and then vaguely about update.
As I keep going back and forth trying to glean the scarcest of information, a picture started forming and I am reminded of a similar situation.
When you run this in debugger and you have just executed the Table I/O, have you ever timed the delay? Once when I did it, it turns out that the delay always fell around two minutes (within 10 secs every time). Once it continued, the Table I/O was always a failure. The delay, then reminded my of a timeout, such as when an operation is waiting for a table to be free and never does become free. My guess is that there are master business functions in use that have Transaction Processing turned on. This will lock up the 'concerned' tables until the End Doc portion of the MBF has completed. As a test, I went to the Begin Doc (and/or the Edit Line) MBF and found the check mark for Transaction Processing and turned it off. Automagically, the process started working. I would guess that this is your same problem.
*******If it is - you then need to REALLY REALLY consider whether turning Transaction Processing off, is the correct solution. Other solutions could exist (i.e. updating using logic encapsulated within a bsfn, or moving the update to after End Doc - also known and an "End Around") which are much preferred over changing the stock code.
One easy test might be to put a Table I/O of a logging table right before the problem Table I/O line and see if it is successful.
Ben Again,