Hi Dave, Here is what I see when running netstat:
root@majestix:~# netstat -nut Active Internet connections (w/o servers) Proto Recv-Q Send-Q Local Address Foreign Address State tcp 0 0 192.168.2.1:35475 192.168.2.100:7147 ESTABLISHED tcp 0 0 134.104.22.254:22 134.104.30.134:49237 ESTABLISHED tcp 0 1 134.104.22.254:42604 198.168.2.100:7147 SYN_SENT The first entry in the list is from ‘telnet 192.168.2.100 7147’ command, while the last is from katcp trying to connect to ROACH2. This times out after a minute or so. It appears like KATCP is using the C-equivalent of “INADDR_ANY” that uses the default interface in the system. Cheers, Ramesh On 14 Jul 2014, at 18:07, David MacMahon <[email protected]> wrote: > Are you sure the katcp python client is using the correct IP? How do you > know it is trying to use eth0 instead of eth1? > > Dave > > On Jul 14, 2014, at 6:24 AM, Ramesh Karuppusamy wrote: > >> >> Hello list, >> >> I have a weird situation with our setup - can someone help me solve this >> issue? >> >> Our head node has two network interfaces - the first on the institute >> network and the interface eth1 is assigned 192.168.2.1. >> >> I assigned a static IP of 192.168.2.100 to our ROACH2 and eth1 above are >> connected to a simple D-Link unmanaged switch. From the head node, I can >> ping ROACH2 or ‘telnet 192.168.2100 7147’ to connect to tcpborphserver. But >> the katcp python client simply doesn't connect to ROACH2, as it tries to use >> the first interface on the head node. Is it possible to force katcp to use >> the eth1 (or ip 192.168.2.1) on the head node? >> >> Thanks! >> >> Cheers, >> Ramesh >
