Unable to install Jetson SDK Components with SDK Manager on Jetson TX2

I’m trying to install Jetson SDK Components on TX2 with SDK Manager,the installation failed and the terminal of SDK Manager shows below.
19:16:37 INFO : CUDA Toolkit for L4T : Unhandled error when running sudo apt-get -y update
19:16:37 INFO : CUDA Toolkit for L4T : exit status 1
19:16:37 INFO : CUDA Toolkit for L4T : [ Package Install Finished with Error ]
19:16:37 INFO : CUDA Toolkit for L4T : [ 92.00 KB used. Disk Avail: 29.31 GB ]
19:16:37 INFO : CUDA Toolkit for L4T : [ NV_L4T_CUDA_TARGET_POST_INSTALL_COMP Install took 5m53s ]
19:16:38 ERROR : CUDA Toolkit for L4T : command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@192.168.1.122:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@192.168.1.122 “~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh” finished with error
19:16:38 INFO : CUDA Toolkit for L4T :
19:16:38 ERROR : CUDA Toolkit for L4T : command terminated with error

I find the possible error info below the “sudo apt-get -y update” command.
19:16:35 INFO : CUDA Toolkit for L4T : E: Failed to fetch http://us.archive.ubuntu.com/ubuntu/dists/xenial/main/binary-arm64/Packages 404 Not Found [IP: 91.189.91.14 80]
19:16:35 INFO : CUDA Toolkit for L4T : E: Failed to fetch http://us.archive.ubuntu.com/ubuntu/dists/xenial-updates/main/binary-arm64/Packages 404 Not Found [IP: 91.189.91.14 80]
19:16:35 INFO : CUDA Toolkit for L4T : E: Some index files failed to download. They have been ignored, or old ones used instead.
How to solve this problem?

If you “ping 91.189.91.14”, do you have a network route to that IP address? The address it is failing to find is the Ubuntu repository for various package install and/or their update. I am guessing that you are behind a proxy or firewall which is blocking the Ubuntu server.

I ping 91.189.91.14,and it seemed to be normal.

nvidia@uavpc:~$ ping 91.189.91.14 -c 4
PING 91.189.91.14 (91.189.91.14) 56(84) bytes of data.
64 bytes from 91.189.91.14: icmp_seq=1 ttl=45 time=247 ms
64 bytes from 91.189.91.14: icmp_seq=2 ttl=45 time=238 ms
64 bytes from 91.189.91.14: icmp_seq=3 ttl=45 time=250 ms
64 bytes from 91.189.91.14: icmp_seq=4 ttl=45 time=245 ms

--- 91.189.91.14 ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3004ms
rtt min/avg/max/mdev = 238.048/245.546/250.593/4.700 ms

Ok, I see it is a case of the server being found, but that the package directory for binary-arm64 is not there. Basically, this particular server supports desktop architectures only.

I don’t know which file or manifest triggers looking here, but it sounds like this failure causes an abort of install. If no abort occurs, then ignoring the error would be ok. In the case where abort is an issue, then it might imply a need to change either the JetPack manifest (which NVIDIA controls), or else to change the Ubuntu default server list (which Ubuntu controls). I just don’t know which manifest refers to this, but if it is the one Ubuntu controls, then there are probably workarounds.

Can you verify this causes an abort of the install? After that, then someone from NVIDIA can check if this is from an Ubuntu file or from an NVIDIA file. If the NVIDIA file, then a direct fix could be provided, and if an Ubuntu file, then a workaround could be provided.

I seem to meet the same problem when I instal CUDA Toolkit for L4T.
But I don’t know how to work it out

The usual first question is if you have a firewall or proxy which might block the address? After that, if no firewall/proxy, can you ping the address and see the server exists? Do logs show “404 Not Found”, or do they show something else?

To help people distinguish between network and package list issues, consider this URL:
http://us.archive.ubuntu.com/ubuntu/dists/xenial/main/binary-arm64/Packages

Even if the arm64 support fails on that server you should expect to be able to reach this URL:
http://us.archive.ubuntu.com/ubuntu/dists/xenial/main/

This URL is based on the stock Ubuntu repository, but different parts of the world may use a slightly different URL. The part which I removed, “binary-arm64/Packages” is the part for support of this specific architecture (a non-PC architecture).

NVIDIA has some control via a manifest file which JetPack/SDK Manager downloads, but Ubuntu controls other parts of this. Someone knowing more about the structure of the manifest for different regions of the world will need to comment, but whoever answers will need to know the above information (it is always good to include a log since URLs and locations in the world may differ for different users).

I used ubuntu16.04 installed on VMware as the host.
I used the original Micro USB cable to connect the host to the target.
Then I checked that the ubuntu16.04 firewall was closed and the Windows firewall was closed.
Then I ping 192.168.55.1, and the result is as follows:

PING 192.168.55.1 (192.168.55.1) 56(84) bytes of data.
64 bytes from 192.168.55.1: icmp_seq=1 ttl=64 time=2.29 ms
64 bytes from 192.168.55.1: icmp_seq=2 ttl=64 time=3.05 ms
64 bytes from 192.168.55.1: icmp_seq=3 ttl=64 time=2.19 ms

It seems to be normal.
I reinstalled it and found the same error.

my error message is as follows:

13:15:35 ERROR : CUDA Toolkit for L4T : command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@192.168.55.1:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@192.168.55.1 “~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh” finished with error
13:15:35 ERROR : CUDA Toolkit for L4T : command terminated with error

I tried to open http://us.archive.ubuntu.com/ubuntu/dists/xenial/main/binary-arm64/Packages,and the page shows “404 Not Found”,and I can get to http://us.archive.ubuntu.com/ubuntu/dists/xenial/main/ normally.I don’t know whether the “sudo apt-get -y update” command cause the installation failure,but I can’t find any error info except that,and the terminal of SDK Manager also shows below:

18:44:10 ERROR : CUDA Toolkit for L4T : command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@192.168.1.122:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@192.168.1.122 "~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh" finished with error
18:44:10 INFO : CUDA Toolkit for L4T :
18:44:10 ERROR : CUDA Toolkit for L4T : command terminated with error
18:44:10 ERROR : install 'CUDA Toolkit for L4T' failure, command < cd '/home/wii/nvidia/nvidia_sdk/JetPack_3.3.1_Linux_GA_P3310' ; '/opt/nvidia/sdkmanager/resources/app/output/installUtils/adapter' -a='install' -c='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' -i='NV_L4T_CUDA_TARGET_POST_INSTALL_COMP' -v='9.0' -l='/home/wii/nvidia/nvidia_sdk/JetPack_3.3.1_Linux_GA_P3310' -L='/home/wii/.nvsdkm/logs/JetPack_3.3.1_Linux_GA' -p='11367' -t='P3310' -e='/home/wii/Downloads/nvidia/sdkm_downloads/cuda-repo-l4t-9-0-local_9.0.252-1_arm64.deb' > terminated with error
18:44:10 ERROR : Unavailable components detected, please check log
18:44:11 ERROR : cuDNN on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
18:44:11 ERROR : TensorRT on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
18:44:11 ERROR : OpenCV on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
18:44:11 ERROR : VisionWorks on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
18:44:11 ERROR : Multimedia API Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,

After you succeed in ping of 192.168.55.1, can you ssh there to your account? Prior to any packages working this would need to succeed, so even if part of the files to download from Ubuntu failed, then some would still work if ssh works; if ssh fails, then it guarantees all packages will fail.

Note that for flash VMs usually fail without work on configuration, but I wouldn’t expect this to be an issue for package installs unless ssh ID is changing in some way.

The next step is to find out if you can manually ssh to your account on that Jetson. If your account name is “ubuntu”, can you do this once ping works:

ssh ubuntu@192.168.55.1

Once we know this works we can move forward. If this does not work, then we have new clues.

I am sure that ssh service is enable. Then I follow your steps and find out the results like this:

ai@ai-virtual-machine:~$ ssh nvidia@192.168.55.1
The authenticity of host '192.168.55.1 (192.168.55.1)' can't be established.
ECDSA key fingerprint is SHA256:5rPfZhqz9qYLMu5g4OwwIxN0xiHTTQfhJgyNoOaUyrE.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added '192.168.55.1' (ECDSA) to the list of known hosts.
nvidia@192.168.55.1's password: 
Welcome to Ubuntu 18.04.3 LTS (GNU/Linux 4.9.140-tegra aarch64)

 * Documentation:  https://help.ubuntu.com
 * Management:     https://landscape.canonical.com
 * Support:        https://ubuntu.com/advantage

This system has been minimized by removing packages and content that are
not required on a system that users do not log into.

To restore this content, you can run the 'unminimize' command.

0 packages can be updated.
0 updates are security updates.

Last login: Tue Nov 26 19:21:48 2019 from 10.92.142.7
nvidia@nvidia-ai:~$ ls
Desktop    Downloads         Music     Public     tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh    Videos
Documents  examples.desktop  Pictures  Templates  tmp_NV_TENSORFLOW_TARGET_POST_INSTALL_COMP.sh
nvidia@nvidia-ai:~$

It makes me confused.

This is a successful ssh login to the “nvidia” account. Presumably you’ve created this on first boot of the Jetson. This ssh login was the first time you logged in to the Jetson from that host PC. So the flash was successful, and in theory, without a VM networking would have worked.

VMs are not supported, but it doesn’t mean you can’t make it work with some effort. VMs must have complete ownership of the USB port no matter how many times it disconnects and reconnects. The virtual wired ethernet which provides the 192.168.55.100 address to the VM must also be permitted, and it is possible that at times this is rejected unless you configure to allow it…but I don’t know for sure.

Because you were able to log in via ssh we can guarantee the flash itself succeeded and the unit is booted and capable of having the extra packages installed via ssh. The VM itself will need internet access and access to the Jetson. The Jetson needs internet access for package updates as well during the extra package addition steps (during flash the Jetson is only accessed over USB without ethernet of any kind, but the host PC does do downloads; after flash, both PC and Jetson access the internet).

I use mobile’s wifi, and I let the PC, the virtual machine system(host), the target to connect to the mobile’s wifi.
The IP address segments are as follows:
the PC:172.20.10.2
the virtual machine system(host):172.20.10.11
the target:172.20.10.10

It failed to work

17:08:29 ERROR : CUDA Toolkit for L4T : command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@10.92.210.136:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@10.92.210.136 "~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh" finished with error
17:08:29 ERROR : CUDA Toolkit for L4T : command terminated with error
17:19:50 ERROR : CUDA Toolkit for L4T : command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@192.168.1.101:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@192.168.1.101 "~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh" finished with error
17:19:50 ERROR : CUDA Toolkit for L4T : command terminated with error
17:52:53 ERROR : CUDA Toolkit for L4T : command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@172.20.10.10:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@172.20.10.10 "~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh" finished with error
17:52:53 ERROR : CUDA Toolkit for L4T : command terminated with error
17:59:38 ERROR : CUDA Toolkit for L4T : command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@172.20.10.10:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@172.20.10.10 "~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh" finished with error
17:59:38 ERROR : CUDA Toolkit for L4T : command terminated with error

I don’t understand why other IP addressesare(like 192.168.1.101 and 10.92.210.136) there.

I think nvidia@172.20.10.10 is the only right way to connect

21:24:35 ERROR : CUDA Toolkit for L4T : command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@192.168.55.1:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@192.168.55.1 "~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh" finished with error
21:24:35 ERROR : CUDA Toolkit for L4T : command terminated with error
21:31:09 ERROR : CUDA Toolkit for L4T : command scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@192.168.55.1:~; ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@192.168.55.1 "~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh" finished with error
21:31:09 ERROR : CUDA Toolkit for L4T : command terminated with error

The above are errors.
one is:

scp -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no /tmp/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh nvidia@192.168.55.1:~

another is:

ssh -t -q -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no nvidia@192.168.55.1 "~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh && rm -f ~/tmp_NV_L4T_CUDA_TARGET_POST_INSTALL_COMP.sh

To understand the error, I decided to run it manually on the host(the virtual machine system).The first command works well.But the second error’s results are as follow:

E: Failed to fetch http://mirrors.tuna.tsinghua.edu.cn/ubuntu/dists/trusty-security/main/binary-arm64/Packages  404  Not Found [IP: 101.6.8.193 80]
E: Failed to fetch http://mirrors.tuna.tsinghua.edu.cn/ubuntu/dists/trusty-updates/main/binary-arm64/Packages  404  Not Found [IP: 101.6.8.193 80]
E: Failed to fetch http://mirrors.tuna.tsinghua.edu.cn/ubuntu/dists/trusty-backports/main/binary-arm64/Packages  404  Not Found [IP: 101.6.8.193 80]
E: Failed to fetch http://mirrors.tuna.tsinghua.edu.cn/ubuntu/dists/trusty-proposed/main/binary-arm64/Packages  404  Not Found [IP: 101.6.8.193 80]
E: Failed to fetch http://mirrors.tuna.tsinghua.edu.cn/ubuntu/dists/trusty/main/binary-arm64/Packages  404  Not Found [IP: 101.6.8.193 80]
E: Some index files failed to download. They have been ignored, or old ones used instead.
Unhandled error when running sudo apt-get -y update

Then I go to http://mirrors.tuna.tsinghua.edu.cn then I find the error.When we follow the path https://mirrors.tuna.tsinghua.edu.cn/ubuntu/dists/trusty-security/main/binary-amd64/, we can’t see the file named “Packages”.In fact, the folder contains the file:

Packages.bz2
Packages.gz
Release

I think it’s the root of the problem.But I don’t know what to do to solve it.
I have tried many other sources, but the situation is the same.
In addition, I can use the same source to update the software on the host(the virtual machine system, ubuntu16.04)

Keep in mind that the 192.168.55.x addresses are the default unless you’ve specified differently. Those addresses are for the micro-B USB cable. Other addresses need to be specified in the SDK Manager when the option to use an alternate address is available. In the case of using WiFi or other addresses instead of the 192.168.55.x over virtual wired USB it will be up to you to make sure the address is valid, and that the host has a route there. Your host address would be in the wrong subnet to talk to 192.168.55.x, but if the Jetson also has the 172.20.10.x network via another mechanism, then this would be valid for cases where you set this as the address to use (after which SDKM should not show any 192.168.55.x log output…instead it would produce 172.20.10.x log output).

Also keep in mind that ssh and scp typically refuse in cases where the address changes unexpectedly. If you’ve manually verified that you can log in via ssh to a given address, then it should be ok for the SDKM to talk to the Jetson. Always manually test ssh login to that address if SDKM gets stuck on software installs.

The web sites you are referring to apparently support desktop PC architectures, but are missing the arm64/aarch64 architecture.

Depending where you are located in the world you may need to expect proxy or firewall issues. I do not know what your manifest file is using, but the missing architecture is usually referring to an official Ubuntu repository which lacks that architecture.

Someone from NVIDIA can probably suggest if this is something the manifest file can be edited for as a workaround, but I personally am uncertain of the specific reason there is a reference to an Ubuntu server without the correct architecture.

Hey,3160100938,I replace my virtual machine with a computer with Ubuntu 16.04 and solve the problem.I think this may help you.

I got some errors while installing CUDA packages:

21:08:24 INFO : CUDA Toolkit for L4T : exec_command [target]:
21:08:24 INFO : CUDA Toolkit for L4T : **********************
21:08:24 INFO : CUDA Toolkit for L4T : #!/bin/bash
21:08:24 INFO : CUDA Toolkit for L4T : set -e
21:08:24 INFO : CUDA Toolkit for L4T : export LC_ALL=en_US.UTF-8 LANG=en_US.UTF-8 LANGUAGE=en_US.UTF-8
21:08:24 INFO : CUDA Toolkit for L4T : function safe_apt_install {
21:08:24 INFO : CUDA Toolkit for L4T : echo “*begin $1, $2”
21:08:24 INFO : CUDA Toolkit for L4T : tmp_log=/tmp/jetpack_dpkg_log.txt
21:08:24 INFO : CUDA Toolkit for L4T : sudo rm -f $tmp_log
21:08:24 INFO : CUDA Toolkit for L4T : set +e; $1 2>&1 | tee tmp_log; CMD_STATUS={PIPESTATUS[0]}; set -e
21:08:24 INFO : CUDA Toolkit for L4T : while [ ${CMD_STATUS} -ne 0 ]; do
21:08:24 INFO : CUDA Toolkit for L4T : if grep -q -e “is locked by another process” -e “E: Could not get lock /var/lib” $tmp_log; then
21:08:24 INFO : CUDA Toolkit for L4T : sudo rm -f $tmp_log
21:08:24 INFO : CUDA Toolkit for L4T : set +e; $1 2>&1 | tee tmp_log; CMD_STATUS={PIPESTATUS[0]}; set -e
21:08:24 INFO : CUDA Toolkit for L4T : echo "Waiting 2 seconds for apt to be available"; sleep {2}
21:08:24 INFO : CUDA Toolkit for L4T : else
21:08:24 INFO : CUDA Toolkit for L4T : echo “Unhandled error when running $1”; exit 1
21:08:24 INFO : CUDA Toolkit for L4T : fi
21:08:24 INFO : CUDA Toolkit for L4T : done
21:08:24 INFO : CUDA Toolkit for L4T : }
21:08:24 INFO : CUDA Toolkit for L4T : safe_apt_install “sudo dpkg -i /opt/nvidia/deb_repos/cuda-repo-l4t-9-0-local_9.0.252-1_arm64.deb” 5
21:08:24 INFO : CUDA Toolkit for L4T : sleep 5
21:08:24 INFO : CUDA Toolkit for L4T : safe_apt_install “sudo apt-key add /var/cuda-repo-9-0-local/7fa2af80.pub” 5
21:08:24 INFO : CUDA Toolkit for L4T : sleep 2
21:08:24 INFO : CUDA Toolkit for L4T : safe_apt_install “sudo apt-get -y update” 5
21:08:24 INFO : CUDA Toolkit for L4T :
21:08:24 INFO : CUDA Toolkit for L4T : safe_apt_install "sudo apt-get -y install cuda-toolkit-9-0 " 5
21:08:24 INFO : CUDA Toolkit for L4T : sudo rm -f /opt/nvidia/deb_repos/cuda-repo-l4t-9-0-local_9.0.252-1_arm64.deb
21:08:24 INFO : CUDA Toolkit for L4T : **********************
21:08:24 INFO : CUDA Toolkit for L4T : [target] nvidia@192.168.55.1’s password:
21:08:24 INFO : CUDA Toolkit for L4T :
21:08:25 INFO : CUDA Toolkit for L4T : [target] nvidia@192.168.55.1’s password:
21:08:25 INFO : CUDA Toolkit for L4T :
21:08:25 INFO : CUDA Toolkit for L4T : *begin sudo dpkg -i /opt/nvidia/deb_repos/cuda-repo-l4t-9-0-local_9.0.252-1_arm64.deb, 5
21:08:25 INFO : CUDA Toolkit for L4T : [target] [sudo] password for nvidia:
21:08:25 INFO : CUDA Toolkit for L4T :
21:08:26 INFO : CUDA Toolkit for L4T : (Reading database … 149390 files and directories currently installed.)
21:08:26 INFO : CUDA Toolkit for L4T : Preparing to unpack …/cuda-repo-l4t-9-0-local_9.0.252-1_arm64.deb …
21:08:26 INFO : CUDA Toolkit for L4T : Unpacking cuda-repo-l4t-9-0-local (9.0.252-1) over (9.0.252-1) …
21:08:33 INFO : CUDA Toolkit for L4T : Setting up cuda-repo-l4t-9-0-local (9.0.252-1) …
21:08:38 INFO : CUDA Toolkit for L4T : *begin sudo apt-key add /var/cuda-repo-9-0-local/7fa2af80.pub, 5
21:08:38 INFO : CUDA Toolkit for L4T : OK
21:08:40 INFO : CUDA Toolkit for L4T : *begin sudo apt-get -y update, 5
21:08:40 INFO : CUDA Toolkit for L4T : Get:1 file:/var/cuda-repo-9-0-local InRelease
21:08:40 INFO : CUDA Toolkit for L4T : Ign:1 file:/var/cuda-repo-9-0-local InRelease
21:08:40 INFO : CUDA Toolkit for L4T : Get:2 file:/var/cuda-repo-9-0-local Release [574 B]
21:08:40 INFO : CUDA Toolkit for L4T : Err:3 http://ports.ubuntu.com/ubuntu-ports xenial InRelease
21:08:40 INFO : CUDA Toolkit for L4T : Temporary failure resolving ‘ports.ubuntu.com
21:08:40 INFO : CUDA Toolkit for L4T : Err:4 http://ports.ubuntu.com/ubuntu-ports xenial-updates InRelease
21:08:40 INFO : CUDA Toolkit for L4T : Temporary failure resolving ‘ports.ubuntu.com
21:08:40 INFO : CUDA Toolkit for L4T : Err:5 http://ports.ubuntu.com/ubuntu-ports xenial-security InRelease
21:08:40 INFO : CUDA Toolkit for L4T : Temporary failure resolving ‘ports.ubuntu.com
21:08:40 INFO : CUDA Toolkit for L4T : Get:2 file:/var/cuda-repo-9-0-local Release [574 B]
21:08:42 INFO : CUDA Toolkit for L4T : Reading package lists…
21:08:42 INFO : CUDA Toolkit for L4T : W: Failed to fetch http://ports.ubuntu.com/ubuntu-ports/dists/xenial/InRelease Temporary failure resolving ‘ports.ubuntu.com
21:08:42 INFO : CUDA Toolkit for L4T : W: Failed to fetch http://ports.ubuntu.com/ubuntu-ports/dists/xenial-updates/InRelease Temporary failure resolving ‘ports.ubuntu.com
21:08:42 INFO : CUDA Toolkit for L4T : W: Failed to fetch http://ports.ubuntu.com/ubuntu-ports/dists/xenial-security/InRelease Temporary failure resolving ‘ports.ubuntu.com
21:08:42 INFO : CUDA Toolkit for L4T : W: Some index files failed to download. They have been ignored, or old ones used instead.
21:08:42 INFO : CUDA Toolkit for L4T : *begin sudo apt-get -y install cuda-toolkit-9-0 , 5
21:08:42 INFO : CUDA Toolkit for L4T : Reading package lists…
21:08:42 INFO : CUDA Toolkit for L4T : Building dependency tree…
21:08:42 INFO : CUDA Toolkit for L4T : Reading state information…
21:08:42 INFO : CUDA Toolkit for L4T : Some packages could not be installed. This may mean that you have
21:08:42 INFO : CUDA Toolkit for L4T : requested an impossible situation or if you are using the unstable
21:08:42 INFO : CUDA Toolkit for L4T : distribution that some required packages have not yet been created
21:08:42 INFO : CUDA Toolkit for L4T : or been moved out of Incoming.
21:08:42 INFO : CUDA Toolkit for L4T : The following information may help to resolve the situation:
21:08:42 INFO : CUDA Toolkit for L4T :
21:08:42 INFO : CUDA Toolkit for L4T : The following packages have unmet dependencies:
21:08:42 INFO : CUDA Toolkit for L4T : cuda-toolkit-9-0 : Depends: cuda-samples-9-0 (>= 9.0.252) but it is not going to be installed
21:08:42 INFO : CUDA Toolkit for L4T : Depends: cuda-documentation-9-0 (>= 9.0.252) but it is not going to be installed
21:08:42 INFO : CUDA Toolkit for L4T : E: Unable to correct problems, you have held broken packages.
21:08:42 INFO : CUDA Toolkit for L4T : Unhandled error when running sudo apt-get -y install cuda-toolkit-9-0
21:08:42 INFO : CUDA Toolkit for L4T : exit status 1
21:08:42 INFO : CUDA Toolkit for L4T : [ Package Install Finished with Error ]
21:08:42 INFO : CUDA Toolkit for L4T : [ 20.00 KB used. Disk Avail: 139.80 GB ]
21:08:42 INFO : CUDA Toolkit for L4T : [ NV_L4T_CUDA_TARGET_POST_INSTALL_COMP Install took 40s ]
21:08:42 ERROR : CUDA Toolkit for L4T : : command safe_apt_install “sudo dpkg -i /opt/nvidia/deb_repos/cuda-repo-l4t-9-0-local_9.0.252-1_arm64.deb” 5; sleep 5; ; safe_apt_install "sudo apt-get -y install cuda-toolkit-9-0 " 5 finished with error
21:08:42 INFO : CUDA Toolkit for L4T :
21:08:42 ERROR : CUDA Toolkit for L4T : command terminated with error
21:08:42 ERROR : CUDA Toolkit for L4T : install ‘CUDA Toolkit for L4T’ failure, command < cd ‘/home/sts04261/nvidia/nvidia_sdk/JetPack_3.3.2_Linux_GA_P3310’ ; ‘/opt/nvidia/sdkmanager/resources/app/output/installUtils/adapter’ -a=‘install’ -c=‘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’ -i=‘NV_L4T_CUDA_TARGET_POST_INSTALL_COMP’ -v=‘9.0’ -l=’/home/sts04261/nvidia/nvidia_sdk/JetPack_3.3.2_Linux_GA_P3310’ -L=’/home/sts04261/.nvsdkm/logs/JetPack_3.3.2_Linux_GA’ -p=‘4647’ -t=‘P3310’ -e=’/home/sts04261/Downloads/nvidia/sdkm_downloads/cuda-repo-l4t-9-0-local_9.0.252-1_arm64.deb’ > terminated with error
21:08:42 ERROR : Unavailable components detected, please check log
21:08:42 ERROR : cuDNN on Target : cuDNN on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
21:08:42 ERROR : TensorRT on Target : TensorRT on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
21:08:42 ERROR : OpenCV on Target : OpenCV on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
21:08:42 ERROR : VisionWorks on Target : VisionWorks on Target Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,
21:08:42 ERROR : Multimedia API : Multimedia API Skipped. Due to its dependencies fail to install: CUDA Toolkit for L4T,

Looks like there is no network to the outside world, or else DNS failed:

21:08:42 INFO : CUDA Toolkit for L4T : W: Failed to fetch http://ports.ubuntu.com/ubuntu-ports/dists/xenial/InRelease Temporary failure resolving 'ports.ubuntu.com'
21:08:42 INFO : CUDA Toolkit for L4T : W: Failed to fetch http://ports.ubuntu.com/ubuntu-ports/dists/xenial-updates/InRelease Temporary failure resolving 'ports.ubuntu.com'
21:08:42 INFO : CUDA Toolkit for L4T : W: Failed to fetch http://ports.ubuntu.com/ubuntu-ports/dists/xenial-security/InRelease Temporary failure resolving 'ports.ubuntu.com'
21:08:42 INFO : CUDA Toolkit for L4T : W: Some index files failed to download. They have been ignored, or old ones used instead.

Do both host and Jetson have network access to the internet?

It worked after connecting to the internet. Thank you.