Re: New details, Re: Troubleshooting Telnet connections that drop for no apparent reason -- HELP?

On 5/17/13 1:02 PM, Dan Kimmel wrote:

Are you replying to the read modified data tags?

Not, so far as we can determine, until after the socket goes EOF. Since there are no fields on the screen, much less any with MDTs set, the response consists of only column, row, and AID bytes, wrapped in a TN5250 packet. And according to the timestamps, it was sent more than a full second *after* the EOF.

--
JHHL

-----Original Message-----

First, at timestamp 16:28:41.519, we get a block of 1297 bytes. This
block renders a rather unremarkable message display screen using a
subfile, using one Write To Display command that renders the screen,
a second WTD that ends immediately after the control characters, and
a Read MDT Fields command.

Then, at timestamp 16:28:42:222 (less than 3/4 of a second later), we
get a failed read on the socket. Specifically, the Java
is.read(byte[] b, int off, int len) on the socket is itself returning
a -1 (which is supposed to indicate an EOF), and no exception is
being thrown.

Then, at timestamp 16:28:43.988 (over 1 3/4 seconds later) we try to
send a 13 byte (15 with the terminating IAC EOR) response.

Since this whole sordid affair has begun, I've been unable to force
an EOF on a Telnet socket with our hardware, not by killing the
connection in WRKTCPSTS, not by varying off the device, and not by
physically unplugging the Ethernet cable.

-- JHHL