problem connecting from NT to a Linux tcp-server

From: Andreas Vierengel (andreas@vierengel.de)
Date: Mon Jun 05 2000 - 06:04:28 EST


Hi,

i've got a strange problem with a windows client...
my server is redhat-linux 6.2...
I have a self written tcp-server program on my internal ip 10.10.5.20 on port
130.
The client (10.20.100.10) is a remote-user with NT...
strace shows, that my server programm is in accept() state.
The client is constantly trying to connect to port 130, but no connection is
established.
I cannot check the client machine (it's not mine) and the user unfortunately
says, that his program works 100%-fine and it must be a bug on my side...
There is an output of tcpdump below.
These 3 lines appear all the time.
Can anybody explain why a connection is not established ?
the client is connected via two cisco-routers, which I cannot check for problems
either :-(

My reading is that the SYN of the client is acked, but why does the client send
an reset afterwards. If I telnet to port 130 from anopther machine in my subnet
the connect will establish...

12:49:10.178303 eth0 < 10.20.100.10.1264 > 10.10.5.20.130: S
932100369:932100369(0) win 8192 <mss 1460> (DF)
12:49:10.178330 eth0 > 10.10.5.20.130 > 10.20.100.10.1264: S
487685310:487685310(0) ack 932100370 win 32120 <mss 1460> (DF)
12:49:10.178553 eth0 < 10.20.100.10.1264 > 10.10.5.20.130: R
932100370:932100370(0) win 0

--Andy
-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to majordomo@vger.rutgers.edu



This archive was generated by hypermail 2b29 : Wed Jun 07 2000 - 21:00:33 EST