Simple Ubuntu Auto-update KILLS Nvidia on Tensorbook

post by marek on Jul 8, 2024

Running Razer x Lambda Tensorbook 2022 on Ubuntu 22.04.1 LTS. After a silent upgrade, well, nvidia-smi returns:

NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.

I ran:

sudo rm -f /etc/apt/sources.list.d/{graphics,nvidia,cuda}* && 
dpkg -l | 
awk '/cuda|lib(accinj64|cu(blas|dart|dnn|fft|inj|pti|rand|solver|sparse)|magma|nccl|npp|nv[^p])|nv(idia|ml)|tensor(flow|board)|torch/ { print $2 }' | 
sudo xargs -or apt -y remove --purge
sudo apt -y update && sudo apt -y install lambda-stack-cuda lambda-tensorbook

per the instructions here How do I reinstall Lambda Stack in Ubuntu on my Tensorbook? | Lambda Docs, and am getting stuck with some kind of driver error.

Starting around here:

DEBUG:Parsing /usr/share/ubuntu-drivers-common/quirks/dell_latitude
Loading new nvidia-550.90.07 DKMS files...
Building for 6.5.0-35-generic
Building for architecture x86_64
Building initial module for 6.5.0-35-generic
ERROR: Cannot create report: [Errno 17] File exists: '/var/crash/nvidia-dkms-550.0.crash'
Error! Bad return status for module build on kernel: 6.5.0-35-generic (x86_64)
Consult /var/lib/dkms/nvidia/550.90.07/build/make.log for more information.
dpkg: error processing package nvidia-dkms-550 (--configure):
 installed nvidia-dkms-550 package post-installation script subprocess returned
error exit status 10

followed by a host of dependency errors:

Setting up python3-torch-cuda (2.3.1+ds-0lambda0.22.04.1) ...
dpkg: dependency problems prevent configuration of nvidia-driver-550:
 nvidia-driver-550 depends on nvidia-dkms-550 (= 550.90.07-0lambda0.22.04.2); ho
wever:
  Package nvidia-dkms-550 is not configured yet.

dpkg: error processing package nvidia-driver-550 (--configure):
 dependency problems - leaving unconfigured
Setting up nvidia-cuda-dev:amd64 (12.2.140~12.2.2-0lambda0.22.04.1) ...
No apport report written because the error message indicates its a followup erro
r from a previous failure.
                          Setting up libxext6:i386 (2:1.3.4-1build1) ...
Setting up libnvidia-fbc1-550:i386 (550.90.07-0lambda0.22.04.2) ...
Setting up nvidia-cuda-toolkit (12.2.140~12.2.2-0lambda0.22.04.1) ...
Setting up nvidia-opencl-dev:amd64 (12.2.140~12.2.2-0lambda0.22.04.1) ...
dpkg: dependency problems prevent configuration of nvidia-driver-535:
 nvidia-driver-535 depends on nvidia-driver-550; however:
  Package nvidia-driver-550 is not configured yet.

dpkg: error processing package nvidia-driver-535 (--configure):
 dependency problems - leaving unconfigured
Setting up libnvidia-gl-550:i386 (550.90.07-0lambda0.22.04.2) ...
No apport report written because the error message indicates its a followup erro
r from a previous failure.
                          Setting up python3-torchvision-cuda (0.18.1-0lambda0.2
2.04.1) ...
Setting up libnvidia-decode-550:i386 (550.90.07-0lambda0.22.04.2) ...
Setting up python3-jaxlib-cuda (0.4.14-0lambda0.22.04.2) ...
Setting up libnvidia-encode-550:i386 (550.90.07-0lambda0.22.04.2) ...
Setting up python3-triton-cuda (2.3.1+llvm18.1.6-0lambda0.22.04.1) ...
dpkg: dependency problems prevent configuration of nvidia-driver-525:
 nvidia-driver-525 depends on nvidia-driver-535; however:
  Package nvidia-driver-535 is not configured yet.

dpkg: error processing package nvidia-driver-525 (--configure):
 dependency problems - leaving unconfigured
Setting up python3-jax-cuda (0.4.14-0lambda0.22.04.1) ...
No apport report written because MaxReports is reached already
                                                              Setting up python3
tensorflow-cuda (2.13.1-0lambda0.22.04.2) ...

Further. the link to the recovery image here is broken lol: Where can I download recovery images? | Lambda Docs.

Is the Tensorbook like a completely dead product? Seems like no official support, all the docs in beta…

post by cody_b on Jul 9, 2024

Further. the link to the recovery image here is broken lol: Where can I download recovery images? | Lambda Docs.

The recovery image can be downloaded from Lambda Stack and recovery images | Lambda Docs.