Good to know it’s not a problem on my end, and yeah I tried removing everything and trying it again but it seems to be just a problem with those specific compressed files.
I also downloaded those files individually from the NVIDIA server and did a bzip2 test on them. They seem to be broken/corrupted at source.
Thanks for the heads up D3_growe. I tried this but it looks like downgrading bzip2 and libz2-1.0 will result in the removal of a bunch of ROS-MELODIC packages for me.
Maybe there’s a different version of the packages that won’t remove ROS-MELODIC? I admit that I wasn’t very careful in which version I selected. Maybe someone here can confirm the best version to install?
Instead of downgrading the system bzip2, the following is a safer alternative (as it doesn’t require running the vulnerable version of bzip2 system-wide). This should also work for @singularity7.
download the .debs, unpack them, and set LD_LIBRARY_PATH and PATH so that that version of bzip2 will be used by sdkmanager:
It appears the internal revision info was not updated in the latest release. Mine looks “old” as well:
tescott@system76-pc:~$ bzip2 --version
bzip2, a block-sorting file compressor. Version 1.0.6, 6-Sept-2010.
Copyright (C) 1996-2010 by Julian Seward.
This program is free software; you can redistribute it and/or modify
it under the terms set out in the LICENSE file, which is included
in the bzip2-1.0.6 source distribution.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
LICENSE file for more details.
I can also confirm that downgrading the bzip2 version got my SDK to progress beyond the failure point at driver install, and it now starts to build the OS image and flash the Xavier module.
If you downgrade and retry it’s going to have the same effect as simply retrying if the download is being corrupted in transit by the CDN. That could explain it working the second time and not the first. Simply, you got a bad download the first time. I can confirm connectivity problems to Nvidia’s servers during the timeframe these prohlems have been happening. Perhaps it’s not the bzip2 version.
I thought it was a corruption problem before investigating the matter more fully. To test that corruption idea I copied archives from two of my colleagues PCs who had not updated their host systems as I had. bzip2 failed to verify those known good archives.
That does sound like bzip2. What’s strange is half the reports are concerning Nvidia tarballs and it’s at the same time as CDN issues, but I guess coincidences happen.