memory leak when using aggregate functions, not fixed since 10.1?
Hello.
Since our switch from 8.1 to 10.2 some years ago, we have an issue which we were able to trivially reproduce, and which is still present up to current 11.2.0.3 version. Attempts to find something similar on metalink failed so far - I haven't seen anything relevant.
Here's the test case. A simple endless pl/sql loop (just as a reproducer) with a single select statement inside:
declare a number;
begin
loop
select min (object_id) into a from user_objects where object_id=0;
end loop;
end;
/
while this loop is running, the session will start eating memory, quickly reaching the point when it gives "unable to allocate PGA memory" error (time depends on the amount of memory available and speed of the cpu, obviously).
Since our switch from 8.1 to 10.2 some years ago, we have an issue which we were able to trivially reproduce, and which is still present up to current 11.2.0.3 version. Attempts to find something similar on metalink failed so far - I haven't seen anything relevant.
Here's the test case. A simple endless pl/sql loop (just as a reproducer) with a single select statement inside:
declare a number;
begin
loop
select min (object_id) into a from user_objects where object_id=0;
end loop;
end;
/
while this loop is running, the session will start eating memory, quickly reaching the point when it gives "unable to allocate PGA memory" error (time depends on the amount of memory available and speed of the cpu, obviously).
0