Hello. I’ve installed Jetpack completely and used TX2 well. However, I accidentally did “chmod 777 -R” at the root of TX2 (I’m not used to LINUX…) and had some problems like CANNOT USE SUDO. So, I’m trying to re-install Jetpack (because I thought flashing would get back the system files of TX2 as original), but the installation doesn’t finish at “determining the IP address of target”. All the conditions are the same as when I succeeded. I’m using a router to connect host to target. Can sudo problem affect the installation also? If so, what can I fix it (including the sudo problem)? Thanks.
Did you successfully flash the first time?
Have you connected the USB cable, and does the initial “flash” bit that overwrites the partitions work?
Some routers filter connections between systems, and don’t let systems on either end talk to the other end.
Also, you can’t update the Jetson with WiFi, it has to be wired Ethernet to work.
Finally, if you hook a screen to the Jetson while you flash, what do you see on the Jetson when it fails?
This is my two cents (for what its worth) for anyone struggling with this issue. Following is my setup.
Failed installation setup
Ubuntu 16.04 LTS VM running on a Late 2017 MacBook Pro (so no USB, just USB-C), ioGear USB-C hub, Apple USB-C to USB and USB to LAN adapter.
I had my Apple USB-C to USB adapter hooked up to my mac and an Apple USB LAN adapter for network connectivity. Both the host machine and the target were connected to a network switch. On the other USB-C port I had my ioGear hub connected to the Jetson via a micro USB. Each time (at least 2) I went through the installation steps meticulously following each step as directed all the way through, put the Jetson in forced recovery. The installer would fail at this step. On one attempt it would flash the OS and in other instances it would not communicate with the target device.
Successful installation setup
Ubuntu 16.04 LTS VM running on a Late 2017 MacBook Pro (so no USB, just USB-C), ioGear USB-C hub. In this setup the network connection for the mac goes through the IOGear hub and the target device is connected to the host via a USB port on the hub.
What I did differently this time in addition to the setup
- I put the target device is forced recovery mode before I attempted to start the installation. On the host (Ubuntu VM) I made sure that the target device (NVIDIA)appeared in the list of USB devices connected to the VM (lsusb on the host).
- I also made sure that I created a virtual bridge on my mac and connected the APX (the target device appears as APX on a mac) is connected to my Ubuntu vm.
Restarted the installation (did not have to put the device in forced recovery since it already is). This time the installation was able to successfully flash the OS image but was again stuck on “determining the IP address of the target device”. I let the target boot up normally.
I unplugged the target from the host machine and restarted the installation this time selecting noinstall for the OS image.Ran ifconfig on the target via ssh to find the IP address of the target device and provided the IP address and the username/password for installation on the target device. Now all tools are installed on the target and I am able to run the samples successfully.