On Sunday 28 June 2009 08:16:40 James Harper wrote:
> > > Okay... I've just had a look through. The GetLastError windows
>
> command
>
> > > is called in three main places. In the win32_tape_read and _write
> > > routines, there is additional parsing of the error done, with
>
> bEO[FTD]
>
> > > set accordingly. Should I leave them as-is and just return
>
> b_errno_win32
>
> > > when the error is unknown (default case) or should I set bEO[FTD]
> > > accordingly and still return b_errno_win32?
> >
> > Unfortunately, I am out *all* day today, and to answer your question
> > completely, I will need to look at the code.  First, I recommend
>
> pointing me
>
> > to the exact line(s) where the problem and bEO[FTD] is happening.
> >
> > If I remember correctly, the the bEO... flag *must* be set correctly
>
> depending
>
> > on the situation.
>
> I think I've got the hang of it.
>
> Unfortunately there are other places that depend on the value of errno,
> in particular EIO for a generic error and ENOSPC for out of space. Just
> setting b_errno_win32 could break those.

Yes, that does not surprise me.  

>
> I think the correct approach is to set simulate EIO when we know what
> the error is and return b_errno_win32 when we don't.
>
> > > The error I am getting is EIO, which is the default case. Hopefully
>
> the
>
> > > proper error reporting will tell me what the actual problem is
>
> now... It
>
> > > might be a few days before I see the error though.
> >
> > OK.
>
> I'm getting errors within the first 30G of the backup so it's pretty
> reproducible within 10 minutes or so. After all that though, the error
> I'm getting is "The request could not be performed because of an I/O
> device error."!

I suspect that your drive is configured to run in asynchronous mode, and to 
work with Bacula it must be in syncronous mode -- i.e. the I/O call should 
not return until it has been properly written to the tape.

>
> Still, all the tests I do on the drive say it's fine so I still think
> the problem is software not hardware although I'm not as sure as I
> was...

Kern

------------------------------------------------------------------------------
_______________________________________________
Bacula-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bacula-devel

Reply via email to