Setting up a processing unit on the local network with remote access privileges to Flow:
Minimum Required Processing Specs
Per 4 Software Instances *1 instance = Traffic, passersby or Demographics | CPU: Intel Core i5 Operating System: Ubuntu OS 20.04 LTS GPU: 8GB RAM: 8GB |
Download the most recent version of Ubuntu
Remote access setup if GUI available
Install TeamViewer: https://www.teamviewer.com/en/download/linux/
Grant “easy access” (Unattended Access) to: support@flow-solutions.com
Configure Client’s firewall to allow TeamViewer to connect out through it:
Allow TCP/UDP Port 5938
Allow TCP Port 443
(Optional) Allow TCP Port 80
(Optional) Add *.teamviewer.com to IP whitelist
Remote access setup if GUI not available
Enable SSH
Set up port forwarding with the port number where requests will be made
Share the machine public IP address here: support@flow-solutions.com
Share user credentials to: support@flow-solutions.com
Installation of Ubuntu and Nvidia Libraries
Before starting make sure that your Graphics card is installed.
Please create a flash USB stick with an Ubuntu image to install it on the machine
When the machine starts, please go to the boot menu and load the disk.
The machine needs to act as a server so make sure that no other OS is installed, you can use the Ubuntu installer to erase the disk and install Ubuntu
While navigating through the Ubuntu installation, you will be asked to install the installation type, please be sure to select the following:
Normal Installation
Install Third-Party Graphics and Media
When the installation is complete please reboot the machine
Open a terminal window in Ubuntu and type the following command:
Sudo -s -> Press Enter and it will prompt for a password, please type the password
Sudo apt-get update -> Press enter and wait for a while
Sudo apt-get upgrade -> Press Enter and wait for a while
Sudo apt-get install cuda-nvidia-toolkit -> Press Enter and follow the instructions
(# The last command can be skipped if you are not comfortable, someone at flow will do it for you.)
Type nvidia-smi
This will allow you to see the processing power of your graphics card
Type nvcc -v
This will allow you to see the CUDA Toolkit version
Need Help?
In case of any problem, feel free to contact us at Flow Solutions, we will assist you with Teamviewer as soon as you send the credentials.