Been reviewing the code again, and I still don't spot anything obvious. Any chance I can get access to the systems were your able to reproduce this on? Perhaps with a couple of extra debugging statements on a few test runs I can figure out whats going on.
Regards, Hiram FuseSource Web: http://fusesource.com/ On Mon, Jan 10, 2011 at 4:33 AM, Lionel Cons <[email protected]> wrote: > Hiram Chirino writes: > > Ok just tried across hosts and still get the same results. You sure > > you running against a recent build? I know there was a bug similar to > > that a little while back. > > Hiram, > > Here is what my colleague (in CC) found: > > | I think the source code which needs attention is: > | > | > http://svn.apache.org/viewvc/activemq/activemq-apollo/trunk/apollo-tcp/src/main/java/org/apache/activemq/apollo/transport/tcp/SslTransport.java > | > | That is in the server's code so probably the apollo developers you are > | in contact with will be able to look at it directly. I believe the > | problem is when the server is sending data as a part of the SSL > | handshake and not all of the data is sent to the channel during a call > | of secure_write() (via that method's call to flush()). I expect the > | apollo developer will be able to follow up on it; if they have any > | queries about what I checked you can put them in contact with me. > > I hope this helps. Feel free to contact David for more information. > > Cheers, > > Lionel >
