October-19th-2020, 09:44 AM
(This post was last modified: October-19th-2020, 01:45 PM by Armin@netPI.)
Well,
my explanation why the TOSIBOX container application works without any routing setup because it is always the initiator of the local TCP/IP traffic sent to the cifx0 interface and its connected devices behind. This is just a standard communication between two local IP addresses where no routing information is needed but just the two IP addresses and no gateway firewall is blocking the access.
In your new use case now the initiator whereas is the PC sitting behind the cifX0. On this PC you need to configure a gateway first of all next to its IP address 192.168.1.x. This gateway needs to be set to the IP address of the cifx0 interface which 192.168.1.111. So all traffic of the PC outside the local network 192.168.1.x will be sent to this gateway address respectively cifx0 instead. As next you should also configure a DNS server on the PC. Usually a DNS server address 8.8.8.8 works fine. Else a simple "ping google.de" will not work on your PC since there is no name resolution configured.
Finally instead of configuring a cifx0 routing path as you did you have to configure the eth0 firewall on your Connect gateway accordingly:
That is all you need to do. This is not a routing problem since the default route is internally configured to eth0 anyway but a firewall problem. You have to trust the eth0 interface as NAT-trusted.
„You never fail until you stop trying.“, Albert Einstein (1879 - 1955)
October-19th-2020, 01:40 PM
Hello Armin,
many thanks for your quick reply.
Yes, it was the NAT.
After enabling NAT support everything is good.
The TOSIBOX is also running.
Greetz from Haan,
Carsten
February-19th-2021, 12:26 PM
(This post was last modified: February-19th-2021, 02:02 PM by Armin@netPI.)
Hi Carsten,
what is the netPI system software version you are running on your new netPI? Make sure you have the version v1.2.2.0 up and running first of all.
You say you can ping 192.168.1.1 which is the address of the local cifX0 ... but how about pinging the remote IP address 192.168.1.100? Does this work also?
Also interesting is which gateway addresses you have configured for eth0 and cifx0 both. It is not allowed for example to have two gateway addresses configured at the same time cause then Linux does not know where route TCP/IP frames if they are not in the pool of local addresses.
Thx
Armin
„You never fail until you stop trying.“, Albert Einstein (1879 - 1955)
February-22nd-2021, 09:28 AM
Hi Carsten,
I feel the firmware update is the most important point here cause we upgraded the cifx0 behaviour specifically in the version 1.2.2.0.
If after the firmware update still no devices connected to any of the dual Ethernet ports "cifx0" are reachable, but you can ping them manually then I remember that there was something that needs to be setup in the TOSIBOX container to enable routing to the cifx0 subnet 192.1681.x ... but for this I am not the expert. It is more than half a year ago I worked with TOSIBOX.
Thx
„You never fail until you stop trying.“, Albert Einstein (1879 - 1955)
February-22nd-2021, 11:20 AM
Well I think your USB stick is not recognized well and hence you netPI isn't updated. From what you are describing everything is fine so far.
You can connect a HDMI cable to netPI's HDMI port an connect a monitor and watch ... there you are able to see if the device makes a normal boot and starts the netPI OS or if it says it is updating the OS while the USB stick is inserted. If the HDMI output says that the netPI is booting normal, the Raspberry Pi CPU was not able to detect the USB stick. The Raspberry CPU is very sensitive with USB sticks an their used USB contoller. It needs a USB controller chip that boot the USB stick in some milliseonds to get recognized.
Maybe you have a second USB stick.
„You never fail until you stop trying.“, Albert Einstein (1879 - 1955)