Re: Error during SQL Server restore fails to propagate back to DBI client on Windows

--- On Mon, 2/22/10, Martin J. Evans <[email protected]> wrote:
> <huger amounts snipped>
> > Yes, it does work for me when {odbc_exec_direct} is
> set to true.  I also understand your explanation of why
> this is the correct way to do things.
> >
> >    
> You need to set odbc_exec_direct - sorry if this prove
> inconvenient for 
> you but there is no other way for DBD::ODBC to know you
> need to do this.
> 
> > The real concern here is why error 3224 seems
> untrappable if this restore is done with {odbc_exec_direct}
> is set to false.  I cannot see how to trap this error
> at all.
> >
> >    
> I don't think there is a way to do this. I replicated your
> situation in 
> C via ODBC and get the same issue. You could perhaps talk
> to Microsoft 
> but I doubt it will get you anywhere.
> 
> > Is it possible to (incorrectly) set odbc_exec_direct
> to false, issue the database restore, and trap error 3224 in
> perl?
> >
> >    
> I don't believe so via Perl and DBD::ODBC or even directly
> via the ODBC API.
> 
> 
> In fairness, perhaps due to my inability to follow where
> you were going 
> I've taken a bit too long to get to the gist of your issues
> and it has 
> arguably highlighted an issue in SQLMoreResults not calling
> set_err in 
> DBI. I always try to help although not always
> successfully.
> 

Thank you very much, Martin!  Your explanation has been tremendously helpful.  It's really appreciated!

If I'm not out of turn, I find this an interesting corner case.  Maybe it merits putting a brief description+caveat in a later version of the DBD::ODBC docs?

Once again, thank you for all of your time and help!

-rt