Hi, Make sure you are using HTTP 1.1 and that u have http-chunking enabled.... When chunking is enabled the internals will just get the stream from the data handler and then will stream it in to the message...
Also in the past I have seen people running in to trouble due to issues with custom stream implementations. One test you can do to figure out that you are not hitting a Axis2 bug or an issue with ur impl is to observe the behaviour when by using a File instead of the stream to create the datahandler.. thanks, Thilina On Thu, Jul 24, 2008 at 3:45 PM, Shalab Goel <[EMAIL PROTECTED]> wrote: > Please help. Any pointers will be appreciated. > > > -----Original Message----- > From: Shalab Goel [mailto:[EMAIL PROTECTED] > Sent: Wednesday, July 23, 2008 1:55 PM > To: [EMAIL PROTECTED]; axis-user@ws.apache.org > Subject: RE: Memory issues with large dynamic generated stream payload > > > Sorry for the partially complete query below, I posted couple of minutes > ago. > > The behavior noted below is seen in both cases -- inbound stream to Axis > Stub layer, and outbound stream from Axis Stub layer. > > With Axis 1.3 and MTOM enabled streaming. > > Regards > Shalab > > -----Original Message----- > From: Shalab Goel [mailto:[EMAIL PROTECTED] > Sent: Wednesday, July 23, 2008 1:32 PM > To: 'axis-user@ws.apache.org' > Subject: Memory issues with large dynamic generated stream payload > > > Hi, > > I am using Axis2 1.3 client to talk with a .NET service. > The request from client layer hands over a stream (wrapped in DataHandler) > to Axis stub layer. > > It seems to the case that when Axis client stub is generating SOAP message, > the memory usage of my client code jumps from 80 MB to 450 MB (or more, > proportional to stream's content size). > > Does Axis Stub layer cache the stream contents in memory before fully > materializing SOAP request? Using Axis 1.3 with MTOM. > > Is this expected behavior? > > If so, is this behavior customizable (via some property) that this in > memory caching of entire stream content not happen. > > If you noted similar issue and have been able to work around it, will look > forward to hearing from you. > > Any suggestions or comments on this thread will be much appreciated. > > Thanks > Shalab > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > -- Thilina Gunarathne - http://thilinag.blogspot.com
