Jetson can not boot with attachment issue. how can i do?

Jetson can not boot with attachment issue. how can i do?
I using one hdmi port (a display), not dual monitor.

Additionally, I’ve been using this Jetson board normally and it lasts just fine. But suddenly the above message appears and jetson is stuck without booting. Please guide me on how to deal with this problem. can you please tell me if you are using custom carrier board or the devkit carrier itself?

Im not sure, our jetson board is the devkit carrier itself.
Our team trying to re-flasing to image on jetson board, but we cant start to default mode, only stay to emergency mode when booting time.
Please, guide to me. How can i do?

If you are flashing the default BSP to the devkit, it should work out of the box. Are you using the SDK manager? If so, try downloading the release 32.7.2 and following the manual steps for flashing the device from the documentation.

Im sorry, i catched our board spec.
This is custom carrier board! So, we can’t apply your basic manual doc. with SDK manager. (image can’t apply with SDK manager)
Would you help me how can our team fix the booting problem?


  • Model : AGX103-11
  • Type : Nvidia Jetson Carrier Board(Rogue-X) by Connect Tech
  • Included
    . AGX Xavier Module Included
    . CTI Active Thermal Installed
    . Wifi/BT Module Installed
    . 1x 1TB SSD Installed

when i trying to re-flashing, jetson do not clear, and stay emergency mode.
Please, see the under figure. and help me.

Just some notes, not an answer…

Tip: If you have serial console, then it is better to get a boot log and dmesg log from this. Screenshots are a bit of a problem.

I noticed in the first screenshot that the i2c is failing for HDMI (the DDC wire is i2c protocol at address 0x50). The video driver would fail any GUI mode. Device tree for your carrier board is probably at fault, and it is the manufacturer of the carrier board which would provide a device tree (perhaps they don’t have one yet for the newer developer preview).

Is the “/home” set up with a separate disk? For example, is it an NVMe or SSD? It seems that a significant part of the issue is that this filesystem is corrupt (issues with device tree probably means you shut down improperly). If so, then you could put the NVMe/SSD on a Linux host PC and check there, and possibly repair. Or you might be stuck formatting and starting this device’s content from scratch.

Your note is the perfect to me! Great thanks!
Our custom board using NVMe 1TB and it has NVMe mount problem in my opinion. When i mount NVMe storage and reboot, then platform can’t mount NVMe. (ubuntu log on failed)

What should I do now? Can only changing to a different NVMe chip solve this problem? or any solution for me?

It looks like the file system on the NVMe is corrupt. Although it is possible this is a hardware issue it is very unlikely to be the cause. Having just the file system error and not some other error says that 99.99% likely the partition was not shut down correctly, and that power loss or other failure caused cached data to fail to be written. Has the system always been properly shut down? If not, then every drive you use will also result in a corrupt file system.

When the partition is so corrupt that it cannot mount it means further write would just cause further corruption. Fixing this is not really a fix as it will cause lost data. The arrangement of nodes has a logic of how they are to interact, which is considered “inconsistent”, and goes beyond just “fixing” it. My thought is that you might want to repair this on another computer, but ultimately, so much is gone that you need to start new and reformat and recreate the content from scratch.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.