tx2 devkit not reachable through netgear fs108p

Hello,

My jetson tx2 devkit flashed with jetpack 28.2.1 is not able to be reached through an unmanaged netgear prosafe fs108p switch.


http://static6.arrow.com/aropdfconversion/e55ff1cab7a0f95ccec7c041af9d6b17d2a94f89/54610-pb00-r.pdf

I get no icmp nor arp packets.

Has anybody experienced the same problem?

Kind regards,

Tanguy

Are you able to log in locally? Either by serial console or keyboard/monitor? If so, what is the output from “ifconfig”? Is that same switch connected directly to a router? Does the router have any kind of security set, e.g., limit to a given MAC?

Hi linuxdev,

I have set a static local network between my laptop, the switch and my jetson. And, yes I am able to log in by serial console.

nvidia@tegra-ubuntu:~$ cat /etc/network/interfaces
# interfaces(5) file used by ifup(8) and ifdown(8)
# Include files from /etc/network/interfaces.d:
#source-directory /etc/network/interfaces.d
auto eth0
#iface eth0 inet dhcp
iface eth0 inet static
        address 192.168.1.3
        netmask 255.255.0.0
nvidia@tegra-ubuntu:~$ ifconfig
eth0      Link encap:Ethernet  HWaddr 00:04:4b:a5:c2:83  
          inet addr:192.168.1.3  Bcast:192.168.255.255  Mask:255.255.0.0
          inet6 addr: fe80::204:4bff:fea5:c283/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:3398230 errors:0 dropped:0 overruns:0 frame:0
          TX packets:22549 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:317990885 (317.9 MB)  TX bytes:3187618 (3.1 MB)
          Interrupt:42 

lo        Link encap:Local Loopback  
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
          RX packets:451 errors:0 dropped:0 overruns:0 frame:0
          TX packets:451 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1 
          RX bytes:31787 (31.7 KB)  TX bytes:31787 (31.7 KB)

It’s not a router, it’s a switch/hub and it’s unmanaged.

Thanks for helping.

What do you see from the “route” command on both the Jetson and on the host (meaning I am assuming it is a Linux PC on that same switch)? What IP address does the host have? Basically I’m trying to find out if the combination of network and route (and indirectly netmask) are valid for the two systems talking to each other without a router.

Yes, it’s an ubuntu 16.04:

user@TaMLap:~$ route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
192.168.1.0     *               255.255.255.0   U     100    0        0 enp0s31f6
nvidia@tegra-ubuntu:~$ route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
link-local      *               255.255.0.0     U     1000   0        0 eth0
192.168.1.0     *               255.255.255.0   U     0      0        0 eth0

The addresses and netmasks and routes are compatible. Nothing basic is in error (although there is something irregular), e.g., collisions.

What is the address of the host system? As long as it isn’t in a different subnet, then it should work.

From your host, what do you see with “traceroute 192.168.1.3”? The address “192.168.1.3” should be the TX2.

The “irregular” note I made is because of this:

iface eth0 inet static
        address 192.168.1.3
        netmask 255.255.<u><b>0</b></u>.0

The netmask should not be “255.255.0.0”…it should be “255.255.255.0”. Even so the route command shows “255.255.255.0”. The private subnet here should be a a “/24”, the “255.255.0.0” is a “/16”. I doubt this would cause the problem unless there is more network setup I’m not seeing. Can you update to “255.255.255.0” and try again?

I’ve set the following config to the TX2 and still no communication:

nvidia@tegra-ubuntu:~$ cat /etc/network/interfaces
# interfaces(5) file used by ifup(8) and ifdown(8)
# Include files from /etc/network/interfaces.d:
#source-directory /etc/network/interfaces.d
auto eth0
#iface eth0 inet dhcp
iface eth0 inet static
	address 192.168.1.2
	netmask 255.255.255.0
	broadcast 192.168.1.255

Same config, same cables on another switch, works directly…

With that current setup, run “dmesg --follow”, and run commands “sudo ifconfig eth0 down” and then “sudo ifconfig eth0 up”. Post the log of what changes in “dmesg --follow” as well as the actual “ifconfig up” and “ifconfig down” command messages.

What is the address and netmask of your laptop host? Assuming it is “eth0” (and it might have a different label, please adjust), what is the output from the laptop for:

ifconfig eth0
route

Here’s the output with “dmesg --follow”

nvidia@tegra-ubuntu:~$ sudo ifdown eth0
RTNETLINK answers: No such process
RTNETLINK answers: Cannot assign requested address
[172227.610943] bcm54xx_low_power_mode(): put phy in iddq-lp mode
nvidia@tegra-ubuntu:~$ sudo ifup eth0
[169965.200586] eqos 2490000.ether_qos eth0: Link is Up - 100Mbps/Full - flow control rx/tx
[169965.210632] __tegra_isomgr_reserve(81920, 0)

Here’s the output of my laptop:

user@TaMLap:~$ ifconfig enp0s31f6
enp0s31f6 Link encap:Ethernet  HWaddr 30:e1:71:cd:53:ec  
          inet addr:192.168.1.1  Bcast:192.168.1.255  Mask:255.255.255.0
          inet6 addr: fe80::a0a7:ac5e:b01b:894c/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:16133498 errors:0 dropped:90070 overruns:0 frame:0
          TX packets:280931 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:10259810776 (10.2 GB)  TX bytes:60691248 (60.6 MB)
          Interrupt:16 Memory:e4500000-e4520000 

user@TaMLap:~$ route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
192.168.1.0     *               255.255.255.0   U     100    0        0 enp0s31f6
nvidia@tegra-ubuntu:~$ ping 192.168.1.1
PING 192.168.1.1 (192.168.1.1) 56(84) bytes of data.
^C
--- 192.168.1.1 ping statistics ---
6 packets transmitted, 0 received, 100% packet loss, time 5040ms

I don’t see an outright error, but something is a bit odd with that. Instead of using “ifup” and “ifdown”, could you try the "sudo ifconfig eth0 "?

FYI, the “100Mbps/Full” is slow due to the switch not supporting gigabit. Other than being slow this shouldn’t matter unless there is some unexpected odd interaction when running at 100Mb/s.

Technically your host can operate without a gateway if the address being communicated with is within the same subnet/netmask combination, and it is…but most systems would have a gateway. If you’ve simply configured this address on your host as a static address without use of a router (and perhaps not internet connected), then this would be a good reason and not an issue. It does make me curious if you could describe if the host is not allowed to talk to the internet? Is the switch not going to an internet capable router? Is there anything else the host can ping?

In order to test the switch, I’ve configured my host to just be able to ping the TX2 through the switch. The switch indeed only allows 10/100Mbps as you can see in the datasheet link I’ve added in my first post. The switch is Auto Uplink and thus MDI-X, I wonder if MDI-X of the switch doesn’t enter in conflict with the TX2’s one. With ethtool I’m not able to disactivate the TX2’s MDI-X, I guess because it’s not supported by the driver.

I see this URL on the topic:
https://techterms.com/definition/mdi

Probably the easiest thing to do is to test with another switch. However, there may be some incompatibility I don’t know about. Normally a regular ethernet cable would do the job to any switch, but I don’t know what might be different about your case. Do you have another switch you can try?

I’ve tested with a D-Link GO-SW-8G and it works directly.

But I need the Netgear to work also. Any known incompatibility between the BCM54610C1IMLG and the RTL8309SB?

I’ve opened a thread on Netgear’s side as well: https://community.netgear.com/t5/Switches-non-manageables/Nvidia-Jetson-TX2-not-reachable-through-fs108p/m-p/1733509#M104

Thanks for helping linuxdev

I finally managed to make it work by limiting the jetson to 10Mbps.

sudo ethtool -s eth0 advertise 0x002