Here is tcpdump log (i try to send 1460 bytes at a time with tcp_write): 11:36:17.891104 IP local_machine > remote_machine: S 6515:6515(0) win 20480 <mss 1460> 11:36:18.124385 IP remote_machine > local_machine: . ack 6763 win 6432 11:36:18.179310 IP local_machine > remote_machine: R 1:1(0) ack 0 win 20480 11:36:21.812658 IP local_machine > remote_machine: S 6515:6515(0) win 20480 <mss 1460> 11:36:22.046403 IP remote_machine > local_machine: S 3771600429:3771600429(0) ack 6516 win 5840 <mss 1460> 11:36:22.193645 IP local_machine > remote_machine: . ack 1 win 20480 11:36:23.375223 IP local_machine > remote_machine: P 1:248(247) ack 1 win 20480 11:36:23.626047 IP remote_machine > local_machine: . ack 248 win 6432 11:36:23.658001 IP local_machine > remote_machine: P 248:1708(1460) ack 1 win 20480 11:36:28.782112 IP local_machine > remote_machine: P 248:1708(1460) ack 1 win 20480 11:36:39.539077 IP local_machine > remote_machine: P 248:1708(1460) ack 1 win 20480 11:36:50.314832 IP local_machine > remote_machine: P 248:1708(1460) ack 1 win 20480 11:36:54.076956 IP local_machine > remote_machine: R 1708:1708(0) ack 1 win 20480 <-- here my software closes connection due to timeout (i set timeout 30 seconds, but i tried 1,2,5 minutes - no difference).
btw. remote_machine is high loaded with different web services running, i browse this services from local_machine with internet explorer without any problems (or pauses). On Tuesday 20 March 2007 18:32, Kieran Mansley wrote: > On Tue, 2007-03-20 at 10:35 +0000, Kieran Mansley wrote: > > On Tue, 2007-03-20 at 12:18 +0200, Vlad wrote: > > > I've solved this by sending data with smaller chunks at a time ( i call > > > tcp_write with TCP_MSS bytes at a time ). unfortunately this slows down > > > uplink. > > > > Hmm, I'm sceptical that this has really solved the problem you described > > I've just realised that Vlad wasn't the original reporter of this > problem, and was just suggesting this as a possible workaround. I'd > assumed he was the same person I'd replied to yesterday - apologies for > getting muddled. However, I still don't think the above is a complete > solution. > > Kieran > > > > _______________________________________________ > lwip-users mailing list > [email protected] > http://lists.nongnu.org/mailman/listinfo/lwip-users _______________________________________________ lwip-users mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/lwip-users
