On Fri, Sep 30, 2011 at 12:32 PM, Tom Sgouros <tomfool@as220.org> wrote:
> Hello all:
> Can anyone tell me what this error implies?
> [rospy.internal][WARNING] 2011-09-30 12:02:32,391: Unknown error initiating
> TCP/IP socket to tahoma:56716 (http://tahoma:32771/): Traceback (most recent
> call last):
>   File
> "/opt/ros/diamondback/stacks/ros_comm/clients/rospy/src/rospy/impl/tcpros_base.py",
> line 474, in connect self.read_header()
>   File
> "/opt/ros/diamondback/stacks/ros_comm/clients/rospy/src/rospy/impl/tcpros_base.py",
> line 526, in read_header
>     self._validate_header(read_ros_handshake_header(self.socket,
> self.read_buff, self.protocol.buff_size))
>   File "/opt/ros/diamondback/ros/core/roslib/src/roslib/network.py", line
> 350, in read_ros_handshake_header
>     d = sock.recv(buff_size)
> error: [Errno 104] Connection reset by peer
> I get this infrequently, but often enough to be a nuisance, say after ten or
> twenty minutes of operation.  The message implies that it's only a warning,
> but one of the nodes on which I see it becomes unaccountably silent when it
> happens.
> Thank you,
>  -Tom

This generally means a connection has been dropped, probably during
reconnect logic.  Thus, that is why one of your nodes is going silent
(i.e. because it is no longer communicating).   Are you communicating
over a wired or wireless link?

FYI: these sort of Q&A posts are best on answers.ros.org, as there are
more people to provide answers there.

 - Ken

> _______________________________________________
> ros-users mailing list
> ros-users@code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>
>