I’m getting this message during updates - any ideas how to open a support case to obtain the key?
Your DGX contract entitles you to Extended Security Maintenance updates for additional packages in the Ubuntu repository. Please contact NVIDIA Support to get your key to enable this capability.
we have two sparks and the cable to join them together but we want to update to latest OS image first.
right now I cant repoduce it the boxes keep reboot with bunch of lines of 00000000’s
it was when I have the cluster connected with amphenol DAC that came with my pair of sparcs it was tossing errors stuck at 13gb transfer speed and telling me peer not enable, seeing mpam_msx_driver_initL bi NSC devices found in firmware: also platform device creation failed -16
CPU’s of 15-19 must have same capacity
gnome keyring daemon
also seeing insuffient power on mlx says 24 watts
Hi, do you have a DGX Spark FE or from an OEM?
nvidia_peermem module is not supported on GB10 units. Did you install a networking driver like DOCA/OFED?
I would recommend reimaging your units which will ensure you are at the latest OS and remove any bad packages
I reimaged them after I wiped the certs and pvst cleared the box still only can get 13gb over dac I used the nccl scripts
But my Pgp keys got wiped box reboots every ten minutes now with oooooooooo
Ooooooooo
Ooooooooo
When it reboots