Lorenz Witte writes:
> Andrew Haley wrote:
> > Lorenz Witte writes:
> > > I have a problem sending strings over RMI. I run a client that passes
> > > over a string via RMI to a server. The server does nothing except
> > > returning the same string. It works fine if the server is run on a Sun
> > > VM. Using JamVM (or Kaffe) with GNU Classpath leads to exceptions on the
> > > client side if the strings are longer than 65535 (ASCII) characters. The
> > > following exception is caused:
> > >
> > > Client exception: java.rmi.MarshalException: error marshalling
> > > arguments; nested exception is:
> > > java.net.SocketException: Broken pipe
> > > java.rmi.MarshalException: error marshalling arguments; nested
exception is:
> > > java.net.SocketException: Broken pipe
> > > at sun.rmi.server.UnicastRef.invoke(UnicastRef.java:129)
> > > at com.actia.rmi.strings.RMIServerImpl_Stub.InOut(Unknown
Source)
> > > at com.actia.rmi.strings.Client.TransmitString(Client.java:45)
> > > at com.actia.rmi.strings.Client.main(Client.java:104)
> > > Caused by: java.net.SocketException: Broken pipe
> > > at java.net.SocketOutputStream.socketWrite0(Native Method)
> > > at
> > > java.net.SocketOutputStream.socketWrite(SocketOutputStream.java:92)
> > > at
java.net.SocketOutputStream.write(SocketOutputStream.java:136)
> > > at
java.io.BufferedOutputStream.write(BufferedOutputStream.java:106)
> > > at
> > > java.io.ObjectOutputStream$BlockDataOutputStream.drain(ObjectOutputStream.java:1639)
> > > at
> > > java.io.ObjectOutputStream$BlockDataOutputStream.writeByte(ObjectOutputStream.java:1677)
> > > at
> > > java.io.ObjectOutputStream.writeFatalException(ObjectOutputStream.java:1354)
> > > at
> > > java.io.ObjectOutputStream.writeObject(ObjectOutputStream.java:281)
> > > at sun.rmi.server.UnicastRef.marshalValue(UnicastRef.java:265)
> >
> > You don't appear to be using GNU Classpath's RMI implementation: this
> > should be gnu.java.rmi.server.UnicastRef.
> Yes, the error occurrs on the client side, i.e. the side where I run a
> Sun JDK. This error, however, only occurs if the server is running GNU
> Classpath.
So it must be our fault, then. :-)
> It seems like a communication error that pops up on the client
> side.
Well, we need to know what's happening. It's hard to be sure with so
little information.
tcpdump might tell us what is going wrong. Alternatively, strace.
Andrew.
Unfortunately I don't have much time to investigate, but I will post an
example code soon (have a load of other stuff to do right now).
Lorenz
