Re: Stale procedures?

In article <usnjc4jo3.fsf_at_yahoo.com>, Galen Boyer <galenboyer_at_yahoo.com> wrote:

> On Thu, 12 Apr 2001, gwestonREMOVE_at_CAPShome.com wrote:
>
> > What we see now is that while the InsertionBottleneck function
> > continues to populate the raw data table, the utility function
> > no longer seems to be working.
>
> I'm a bit unclear.
>
> ProcA executes Func1, Func2, Func3.
> Func2 executes Func2.1, Func2.2, Func2.3
>
> UserA is owner of all objects
>
> UserB execute ProcA
>
> UserA fixes Func2 and then recompiles Func2.1, 2.2 and 2.3
> because they are invalid?
>
> Is this the scenario?

Close, but not quite. The only modification I'd make to your summary is that UserA fixes Func2.2 (which causes Func2 to go invalid which causes ProcA to go invalid) and then recompiles all invalid objects in the UserA schema. During this event, UserB has sat idle but connected. The issue is that when UserB starts working again, ProcA appears to execute fine (the raw data table gets populated and output arguments that are only set at the end have provably correct values in them) but Func2 (which reads a row from the raw data table and populates other tables) seems to have been bypassed silently.

G Received on Sat Apr 14 2001 - 00:20:59 CEST