"Tom Potts" <anonymous@discussions.microsoft.com> wrote in message
news:006b01c3997a$bf98fc40$a301280a@phx.gbl...
> I have a routine that uses xp_readmail to read one mail at
> a time form an outlook mailbox profile and then process'
> the saved files.
> The process has to be 'broken' for sensible error handling.
> The process works fine on a non clustered machine as it
> writes to "c:\documents~1\temp\newfilename.ext" - this is
> then stored in the db for later work.
>
> On a clustered machine however it can be picked up by the
> other machines in the cluster and of course the filepath
> is invalid!
>
> I suppose what I'm asking is how can one set the 'temp'
> directory of a mail profile?
>

>-----Original Message-----
>I can't answer the question "how can one set the 'temp'
directory of a mail
>profile?" But if that's something you can configure, you
could try to place
>it on a file share resource so that it's always visible
to xp_sendmail no
>matter pn which node it's running, and you have only a
single temp directory
>for the mail profile. I must admit, though, I've never
tried this with a
>MAPI profile. I'm staying away from MAPI/Outlook on my
SQL clusters as far
>as I can.
>
>--
>Linchi Shea
>linchi_shea@NOSPAMml.com
>
>
>"Tom Potts" <anonymous@discussions.microsoft.com> wrote
in message
>news:006b01c3997a$bf98fc40$a301280a@phx.gbl...
>> I have a routine that uses xp_readmail to read one mail
at
>> a time form an outlook mailbox profile and then process'
>> the saved files.
>> The process has to be 'broken' for sensible error
handling.
>> The process works fine on a non clustered machine as it
>> writes to "c:\documents~1\temp\newfilename.ext" - this
is
>> then stored in the db for later work.
>>
>> On a clustered machine however it can be picked up by
the
>> other machines in the cluster and of course the filepath
>> is invalid!
>>
>> I suppose what I'm asking is how can one set the 'temp'
>> directory of a mail profile?
>>
>
>
>.