Sdkmanager install drive software 10.0 issue

I still saw the same errors in your logs.
Before running SDK Manager to install, please fix them and make sure “sudo apt-get update” has no problem.

@VickNV I have fixed the other issues except this:
W: Skipping acquire of configured file ‘main/i18n/Translation-en_US’ as repository ‘https://cfengine-package-repos.s3.amazonaws.com/pub/apt/packages stable InRelease’ doesn’t have the component ‘main’ (component misspelt in sources.list?)
W: Skipping acquire of configured file ‘main/i18n/Translation-en’ as repository ‘https://cfengine-package-repos.s3.amazonaws.com/pub/apt/packages stable InRelease’ doesn’t have the component ‘main’ (component misspelt in sources.list?)
W: Skipping acquire of configured file ‘main/dep11/Components-amd64.yml’ as repository ‘https://cfengine-package-repos.s3.amazonaws.com/pub/apt/packages stable InRelease’ doesn’t have the component ‘main’ (component misspelt in sources.list?)
W: Skipping acquire of configured file ‘main/dep11/icons-48x48.tar’ as repository ‘https://cfengine-package-repos.s3.amazonaws.com/pub/apt/packages stable InRelease’ doesn’t have the component ‘main’ (component misspelt in sources.list?)
W: Skipping acquire of configured file ‘main/dep11/icons-64x64.tar’ as repository ‘https://cfengine-package-repos.s3.amazonaws.com/pub/apt/packages stable InRelease’ doesn’t have the component ‘main’ (component misspelt in sources.list?)
W: Skipping acquire of configured file ‘main/dep11/icons-64x64@2.tar’ as repository ‘https://cfengine-package-repos.s3.amazonaws.com/pub/apt/packages stable InRelease’ doesn’t have the component ‘main’ (component misspelt in sources.list?)
W: Skipping acquire of configured file ‘main/dep11/icons-128x128.tar’ as repository ‘https://cfengine-package-repos.s3.amazonaws.com/pub/apt/packages stable InRelease’ doesn’t have the component ‘main’ (component misspelt in sources.list?)
W: Skipping acquire of configured file ‘main/cnf/Commands-amd64’ as repository ‘https://cfengine-package-repos.s3.amazonaws.com/pub/apt/packages stable InRelease’ doesn’t have the component ‘main’ (component misspelt in sources.list?)

Then try if the installation has any problem.

@VickNV tried with errors. The amazonaws error has to be fixed since it prevents the installation:
2:34:07 ERROR: OSS and NV packages - target_image: W: Skipping acquire of configured file ‘main/i18n/Translation-en’ as repository ‘https://cfengine-package-repos.s3.amazonaws.com/pub/apt/packages stable InRelease’ doesn’t have the component ‘main’ (component misspelt in sources.list?)

12:34:18 ERROR: OSS and NV packages - target_image: E: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages.

12:34:18 ERROR: OSS and NV packages - target_image: ERROR: Installation failed.

12:34:18 ERROR: OSS and NV packages - target_image: [error]: [exit_code]: 1: [exec_command]: /bin/bash -c /home/shangping.guo/Downloads/nvidia/sdkm_downloads/drive-t186ref-linux-5.1.6.1-16911241-oss-minimal-sdk.run --nox11 | tee -a /home/shangping.guo/.nvsdkm/logs/DRIVE_Software_10.0_Linux_OS/NV_OSS_AND_NV_PACKAGES_SDK_COMP.log; exit ${PIPESTATUS[${#PIPESTATUS[@]}-2]}; [error_message]: ERROR: Installation failed.

12:34:18 ERROR: OSS and NV packages - target_image: command terminated with error

12:34:18 SUMMARY: OSS and NV packages - target_image: First Error: Installation failed.

12:34:18 SUMMARY: Flash Xavier A+B in parallel - flash: Depends on failed component

12:34:18 SUMMARY: Drive Platform Targetfs Patch - target_image: Depends on failed component

12:34:18 SUMMARY: Drive Platform Targetfs - target_image: Depends on failed component

12:34:18 SUMMARY: HYPERION-7.1 - target_image: Depends on failed component

12:34:18 SUMMARY: DDS Framework - target_image: Depends on failed component

12:34:18 SUMMARY: TensorRT (with cuDNN 7.5) on Target - target_image: Depends on failed component

12:34:18 SUMMARY: CUDA on Target - target_image: Depends on failed component

12:34:18 SUMMARY: Aurix Updater - target_image: Depends on failed component

12:34:18 SUMMARY: DRIVE AGX Packages - target_image: Depends on failed component

Is it in your /etc/apt/sources.list? Maybe you can remove it.

@VickNV after the removal, I got the following error:
13:04:45 ERROR: OSS and NV packages - target_image: E: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages.

13:04:45 ERROR: OSS and NV packages - target_image: ERROR: Installation failed.

13:04:45 ERROR: OSS and NV packages - target_image: [error]: [exit_code]: 1: [exec_command]: /bin/bash -c /home/shangping.guo/Downloads/nvidia/sdkm_downloads/drive-t186ref-linux-5.1.6.1-16911241-oss-minimal-sdk.run --nox11 | tee -a /home/shangping.guo/.nvsdkm/logs/DRIVE_Software_10.0_Linux_OS/NV_OSS_AND_NV_PACKAGES_SDK_COMP.log; exit ${PIPESTATUS[${#PIPESTATUS[@]}-2]}; [error_message]: ERROR: Installation failed.

13:04:45 ERROR: OSS and NV packages - target_image: command terminated with error

13:04:45 SUMMARY: OSS and NV packages - target_image: First Error: Installation failed.

13:04:45 SUMMARY: Flash Xavier A+B in parallel - flash: Depends on failed component

13:04:45 SUMMARY: Drive Platform Targetfs Patch - target_image: Depends on failed component

13:04:45 SUMMARY: Drive Platform Targetfs - target_image: Depends on failed component

13:04:45 SUMMARY: HYPERION-7.1 - target_image: Depends on failed component

13:04:45 SUMMARY: DDS Framework - target_image: Depends on failed component

13:04:45 SUMMARY: TensorRT (with cuDNN 7.5) on Target - target_image: Depends on failed component

13:04:45 SUMMARY: CUDA on Target - target_image: Depends on failed component

13:04:45 SUMMARY: Aurix Updater - target_image: Depends on failed component

13:04:45 SUMMARY: DRIVE AGX Packages - target_image: Depends on failed component

@VickNV I have another question, I need install qnx sdp on pegasus, I am confused on the drive qnx sdk, how can I get it installed? Shall I open a new issue for this?

Please provide current ~/.nvsdkm logs.

sdklog.tar.gz (2.3 MB)
@VickNV

It looks your system was pretty messed up.
Below are your current problems on linux-libc-dev:i386 installation. You need to fix them first.
But I would really suggest you use a clean ubuntu 18.04 system to save the time.

13:04:44.930 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: Non interactive : Installing package linux-libc-dev:i386
13:04:45.099 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: Reading package lists…
13:04:45.357 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: Building dependency tree…
13:04:45.365 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: Reading state information…
13:04:45.508 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: Some packages could not be installed. This may mean that you have
13:04:45.508 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: requested an impossible situation or if you are using the unstable
13:04:45.508 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: distribution that some required packages have not yet been created
13:04:45.508 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: or been moved out of Incoming.
13:04:45.509 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: The following information may help to resolve the situation:
13:04:45.509 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX:
13:04:45.509 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: The following packages have unmet dependencies:
13:04:45.593 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: libboost1.65-dev : Depends: libstdc+±4.8-dev but it is not going to be installed or
13:04:45.593 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: libstdc+±dev
13:04:45.598 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: libopenmpi-dev : Depends: libhwloc-dev but it is not going to be installed
13:04:45.673 - error: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: E: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages.
13:04:45.673 - info: Event: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX - error is: E: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages.
13:04:45.855 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX:
13:04:45.855 - info: NV_OSS_AND_NV_PACKAGES_SDK_COMP@DDPX: ERROR: linux-libc-dev:i386 failed to install

@VickNV On another linux machine, I tried it, the host is not installed successfully, I did not flash the target, but the machine connects to pegasus successfully. On this linux machine, there is no firewall-cmd installed, so I think that is the reason. The firewall-cmd may block the connection for data flow. But right now, it reports error: pdk jason file not selected. What is that?

Thanks for letting us know.
Please create another topic for the issue and attach a file with all logs under ~/.nvsdkm. Thanks.

@VickNV I think we are on the wrong track. My purpose is to do nsight profiling on drive os 5.1. It seems that the 2019.3 and the 2021.1 both have the same problem failing on cuda injection initialization. I think the problem is on the target machine, and we do not need to try to fix the 2019 nsight problem which involves too many client-server problems. I think we shall focus on the why the cuda injection failure. If that is the case, we can close this issue and focus on that problem.

Did you reply to the wrong topic? This is for DRIVE Software 10.0 installation issue.

@VickNV I mean we can close this issue if it does not relate to the profiling. We do not need the host to do the profiling. Using 2021.3 on the target we can do the profiling too

Sorry, I lost.
This topic is to make you reflash your target system successfully.
But if you think don’t need it to clarify the night issue now, we can always come back later when necessary.

BTW, please always create a new topic for a new issue. It will be clearer. Thanks.

@VickNV So you still think that the problem is the target system? We have several 5.1 and several 5.2, all 5.1 does not work, but all 5.2 works with the same host. I do not think that all the 5.1 system is not right. What is your idea?

I didn’t know you tried with DRIVE OS 5.2 and it was working.
Please share the details in Nsys cannot collect cuda information on Drive OS 5.1 (if this is still an issue, we should discuss there).

@VickNV OK, I will do it after the meeting with nvidia representative on the qnx issue. Thanks