Re: [lwip-users] Re-Transmission from PC is ignored due to sequence number

[Patrick Klos]

On 4/23/2018 1:32 AM, Daniel L wrote:
> Simon Goldschmidt wrote
> > How come your PC retransmits with the same sequence number after 300ms?
> > The
> > Modbus implementations I know are retransmitting the application data
> > (i.e.
> > Modbus frame), not the TCP seqnos.
> The retransmission is not done by the Modbus implementation but by the
> Windows 7 TCP/IP stack.
> For me this seems OK. Because the initial modbus frame from PC is not
> acknowledged by our device (the frame containing the ACK is the frame
> getting lost), Windows has to do a retransmission after some time.
> Do you see a problem here?
>
> Thinking about the retransmission from PC side:
> Shouldn't lwip also do a retransmission of the lost frame?
> You know, these bytes never get acknowledged by PC.

Yes, the LwIP side should have retransmitted the packet.  Packet 21 of 
the trace file shows an ACK of 00009b45, so even though the LwIP device 
thought it sent a response to the PC (bumping the ACK to 00009b56), LwIP 
should have noticed that and resent the data from the previous packet.

> Patrick Klos-2 wrote
> > Based on your description, yes, LwIP is acting properly.  You need to
> > find out why the payload packet isn't making it to the PC?  How big is
> > the payload packet?  Can you share a file with the packets from Wireshark?
> The frame just contains a dozen bytes of TCP-payload, thus, with all the
> TCP/IP headers the raw ethernet frame is about 65 bytes.
> We are investigating where the frame gets lost.
> This may be some hardware issue.
>
> However, as Simon Goldschmidt wrote, TCP is a reliable protocol.
> As far as my understanding goes, one lost ethernet frame should be no
> problem because TCP/IP has mechanisms to handle this.
> Again I'm wondering why lwip does not perform a retransmission after not
> getting acknowledged by PC-side?
>
> trace.pcap <http://lwip.100.n7.nabble.com/file/t2047/trace.pcap>

Looking at the trace file and re-reading your symptoms, I wonder if the 
ethernet transmitter on your embedded device failed to send that 
specific payload packet?  Does your ethernet controller have a status 
register that might indicate if an error occurred during transmission?

Either that, or the packet got lost between the embedded device and the 
Windows PC?  Is there a switch or other network device between the 
device and the PC?  You might need to tap at the embedded device itself 
to see for sure what packets are actually leaving the device.  I use a 
Netgear GS108E (with port mirroring) for such a setup so I can 
monitor/verify the packets on the embedded system's link.

Still, it's unclear why the LwIP stack sent packet 22 with the SEQuence 
number 0009b56 when that data has not been ACKed yet?

Patrick