As I understand it, there are two socket reset conditions:

1) ECONNRESET

Connection reset by peer. Verify destination communication software is
active and ready to accept connections before proceeding.

2) ENETRESET

Remote host dropped network communications. Verify the communication
software on the remote host you were communicating with is available and
ready to accept connections.

The above was cut & paste from an IBM diagnostics web page.

If the Sun code is anything like the GNU code, the JVM is just reporting an
ENETRESET error bubbling up from the socket layer (q.v.,
http://gcc.gnu.org/ml/java-patches/2001-q3/msg00023.html).

FWIW, the earlier bug I was recalling had to do with connection timeouts,
which was a bug in the Windows JVM.  That bug was fixed by the person adding
an explicit         socket.setSoTimeout( timeout ); statement to the handler
code, but I don't bebelive that it has been committed to the CVS.

        --- Noel


--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to