Torstein,

Applied your patch to CVS. Question: For your other problem, can you switch to 
CommonsHTTPSender
and see if you can reproduce the problem? 

Thanks,
dims

--- Torstein Olsen <[EMAIL PROTECTED]> wrote:
> Patch under bugzilla #25385.
> 
> On Tue, 2003-12-09 at 18:55, Davanum Srinivas wrote:
> > Can you please send us the patch for "fixed a similar error in the 
> > ManagedMemoryDataSource"?
> > Usually we use bugzilla to report problems so that we don't lose track of it. 
> > 
> > http://ws.apache.org/axis/bugs.html
> > http://nagoya.apache.org/wiki/apachewiki.cgi?AxisProjectPages/SubmitPatches
> > 
> > Thanks,
> > dims
> > 
> > --- Torstein Olsen <[EMAIL PROTECTED]> wrote:
> > > We have a curious problem that we can reproduce and pinpoint where it
> > > happens, but haven't been able to fix.
> > > 
> > > We try to use Axis to transfer quite large files as attachments > 2Gb.
> > > 
> > > When the transfer is nearly complete both client and server hang.
> > > 
> > > We have narrowed the problem down to the following areas:
> > > 
> > > Server:
> > >   In class "org.apache.axis.attachments.BoundaryDelimitedStream" the
> > > server hangs on the line (266):
> > > 
> > >   br = is.read(b, brTotal + start, length - brTotal);
> > > 
> > > in the method "ReadFromStream"
> > > 
> > > Client:
> > >   In class "org.apache.axis.transport.http.HTTPSender" the client hangs
> > > on the line (508):
> > > 
> > >   b = (byte) inp.read();
> > > 
> > > in the method "readHeadersFromSocket"
> > > 
> > > When we look through the code it looks as though the client has finished
> > > writing the data to the outputstream and waits for a reply from the
> > > server. 
> > > 
> > > At the same time the servers hasn't received the last block of the file
> > > and waits for "eos".
> > > 
> > > This only occurs when the attachment is over 2Gb which leads us to a
> > > possible conclision that there is and int value involved somewhere that
> > > should have been a long.
> > > 
> > > We have found and fixed a similar error in the ManagedMemoryDataSource
> > > class, but obviously there is at least one more.
> > > 
> > > Any suggestions as to where we should look will be appreciated.
> > > 
> > > tia
> > > -- 
> > > Torstein Olsen <[EMAIL PROTECTED]>
> > > NB
> > 
> > 
> > =====
> > Davanum Srinivas - http://webservices.apache.org/~dims/
> brain, v: [as in "to brain"] To rebuke bluntly, but not pointedly; to
> dispel a source of error in an opponent. -- Ambrose Bierce, "The Devil's
> Dictionary"


=====
Davanum Srinivas - http://webservices.apache.org/~dims/

Reply via email to