YESSS! That is precisely the scenario I described as "distributed
queries causing locks".
On 10/29/18 11:06 AM, Mark W. Farnham wrote:
--
When a transaction is interrupted by any sort of a network hiccup (most insidiously between a client machine and the middle tier with no loss of connectivity between a middle tier and the database apparent unless a timeout is configured) then the locks on the statements of the transaction thus far in play can be held for a very long time.