[jcifs] NoRouteToHost persists
Michael B Allen
mballen at erols.com
Sat Feb 16 18:04:25 EST 2002
On Fri, 15 Feb 2002 22:34:05 -0800
rjw <rob at wygand.com> wrote:
> Seeing some bad behavior. Basically, when I get a no route to host
> exception, it seems like jCIFS gets stuck... any subsequent attempts to
> contact that host a fail. I imagine we're caching something we should
> not be. Note that to get the no route to host I simply pulled the
> network cable, then later plugged it back in.
In SmbTransport an IOException such as NoRouteToHost should cause
tryClose( true ). If you see "exception reading from socket input"
then tryClose was ideed called and the socket should be shutdown. The
SmbTransport object is technically still "cached" in the connections
Vector but it's socket is null at that point so jCIFS isn't really
caching anything of interest here.
Can you verify that this stuff is happening like I say it is? Make your
own log mask and use Log.setPrintWriter to write the log messages to a
file. Then place some debugging messages in and around tryClose to try
and see what's going on.
I fear this may just be something buried in the JVM socket layer though
because at first glance I don't really see how jCIFS could be screwing
May The Source be with you.
More information about the jcifs