RTE3 Use real time ethernet as "normal" ethernet interface - Printable Version +- Hilscher Community Forum (https://forum.hilscher.com) +-- Forum: netPI 3 - Docker featuring Industrial Raspberry Pi 3 platform (https://forum.hilscher.com/forum-1.html) +--- Forum: Software (https://forum.hilscher.com/forum-5.html) +--- Thread: RTE3 Use real time ethernet as "normal" ethernet interface (/thread-410.html) |
RE: RTE3 Use real time ethernet as "normal" ethernet interface - Armin@netPI - April-15th-2019 When I am back to the office and able to test I will add the following three lines to support the routing as well: Code: NETWORK=$((i1 & m1)).$((i2 & m2)).$((i3 & m3)).$((i4 & m4)) RE: RTE3 Use real time ethernet as "normal" ethernet interface - Andi - April-16th-2019 Thanks, now the IP-Adress and Subnet-Mask are working correctly with the posted configuration. BR, Andi RE: RTE3 Use real time ethernet as "normal" ethernet interface - Andi - April-18th-2019 Hey Armin, have you tried to add the routing already? Thanks, Andi RE: RTE3 Use real time ethernet as "normal" ethernet interface - Armin@netPI - April-18th-2019 No, will do it after my vacation next week. I have no test system with me and don’t want to hack in untested stuff. Have you tried the three commands I sent as code extract and can you share your experiences with it? Then the implementation has been verified by an independent person RE: RTE3 Use real time ethernet as "normal" ethernet interface - Armin@netPI - April-20th-2019 Routing is now implemented. RE: RTE3 Use real time ethernet as "normal" ethernet interface - Andi - April-30th-2019 Hey Armin, sorry for the late response, haven't been in office. The routing talbe still looks the same as before: Code: root@2fc071664fb9:/# ip route Maybe it's because of the errors in the log-file: Code: starting ssh ..., Any idea what went wrong? Still using the same configuration as written before. Thanks, Andi RE: RTE3 Use real time ethernet as "normal" ethernet interface - Armin@netPI - April-30th-2019 Hello Andi, I see the same effect like you do. I tried to understand the root cause of the error "RTNETLINK answers: Network is down" and I have learned that using an "ip route" command on any interface needs to have it in "up" state before the command can be used. This is not how it is realized today. I never saw this problem cause my interface was "up" always and hence all "ip route" commands worked right away. Meanwhile I am getting this in my container when I am using your IP setup Code: root@b8e92e0e5baf:/# ip route Is it how it shall look like? RE: RTE3 Use real time ethernet as "normal" ethernet interface - Armin@netPI - May-2nd-2019 Container updated Getting this output now Code: root@e7dec71e8429:/# ip route Thx Armin RE: RTE3 Use real time ethernet as "normal" ethernet interface - Andi - May-2nd-2019 Hey, I get the same output now. The Problem that now occurs is that the "bridged" network mode is not working. the IP-Adresses look like this: Code: 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1 As you can see, the eth0 is using an IP-Adress, which is located within the subnet of the Network-Bridge. The docker-network-bridge than connects this IP-Adress to the host's IP-Adress of this adapter. The cifx0 adapter directly gets the IP-Adress, which shall be used for remote communication and not that of the internal Network-Bridge, which is right? If I connect the netPi over a switch with another device within the same subnet, I cannot ping either the device from within netPi's container, nor the netPi from the device, using 10.7.1.115 as IP-Adress. Could it be that docker is somehow prohibiting the communication or that it gets stuck within the container? Have you been able to ping another device using the cifx0-adapter? Thanks, Andi RE: RTE3 Use real time ethernet as "normal" ethernet interface - Armin@netPI - May-2nd-2019 Well Andi, you configured the gateway 10.7.1.1 and this is how it is configured now. So once you call just "ping 10.7.1.20" for example within the container, the ethernet protocol will send an ARP telegram from 10.7.1.115 to the IP address 10.7.1.1 to ask for the MAC address of 10.7.1.1 first of all. And if the 10.7.1.1 device does not exits, then the ping command will not get back any response and hence will fail. Are you sure you have a gatewas running at 10.7.1.1? Thx Armin |