feat: update driver info and hpc computing

docs/pages/Gitlab-CI.md

@@ -9,6 +9,10 @@ Most Gitlab CI jobs are running whith each push to the repository. Additionally,
 
 Almost all jobs are based on the Docker image from the Gitlab CI Docker Registry.
 
+## Phoenix HPC Usage
+The institute uses a "Funktionsaccount" to run simulations from the CI-Pipeline on the Phoenix HPC cluster (for example the multi-gpu test).
+The CI job uses the [HPC-Rocket](https://github.com/SvenMarcus/hpc-rocket) to communicate with the Phoenix HPC cluster. In order to access the cluster, some environments-variables, like the hostname, username and ssh-key needs to be provided. The environment-variables are stored in the [Gitlab CI/CD settings](https://git.rz.tu-bs.de/irmb/VirtualFluids/-/settings/ci_cd) of the project.
+
 ## Gitlab Runner
 
 There a currently three machines, where the ci jobs are running. The machines are called `gitlab-runner01`, `gitlab-runner02` and `gitlab-runner03`. The machines are running on the Gitlab Runner software.
@@ -17,6 +21,8 @@ The runner are registered in the Gitlab project and are using the tags `linux`,
 All Gitlab runner are shared to all Maintainer in this sub-project: https://git.rz.tu-bs.de/irmb/shared/runner/-/settings/ci_cd.
 With this they can use the runner in their own projects. They need to be enabled for every new project. (e.g. virtualfluids project: https://git.rz.tu-bs.de/irmb/VirtualFluids/-/settings/ci_cd).
 
+The Gitlab-Runner are registered to the GITZ Gitlab and can be found in the CI/CD - Runner part of the project settings.
+
 
 ### gitlab-runner01 and gitlab-runner02 (Linux)
 - IP: 134.169.81.206/207
@@ -26,8 +32,10 @@ With this they can use the runner in their own projects. They need to be enabled
 
 
 
-#### Update NVIDIA Driver:
-0. Download a newer driver (optional). Only on demand, for example a new cuda version requires a new driver. Its needs to be checked if the new driver is compatible with the GPU.
+#### Update / Install NVIDIA Driver:
+Nvidia Driver needs to be reinstalled after every kernel update of the OS.
+
+0. (Optional!) Download a newer driver. Only on demand, for example a new cuda version requires a new driver. Its needs to be checked if the new driver is compatible with the GPU. With a new driver the installation script might needs to be updated - so start the installation with a new driver manually first.
     - get newest driver link from here: https://www.nvidia.com/de-de/drivers/
     - copy the download Link. e.g. https://us.download.nvidia.com/XFree86/Linux-x86_64/550.90.07/NVIDIA-Linux-x86_64-550.90.07.run
     - login to runner via ssh
@@ -41,7 +49,7 @@ With this they can use the runner in their own projects. They need to be enabled
     ```
 
 1. install driver ( e.g. 550.90.07.):
-Pass the driver file name as argument to the script.
+Pass the driver file name as argument to the script (Attention: with a new driver version the installation script might needs to be updated).
 ```
 sudo ./install_nvidia_driver.sh NVIDIA-Linux-x86_64-550.90.07.run
 ```
@@ -73,6 +81,12 @@ Fri Jun 14 08:44:18 2024
 +-----------------------------------------------------------------------------------------+
 ```
 
+#### Clean old docker images
+Regularly clean old dangling docker images to free up disk space.
+```
+docker system prune -a
+```
+
 ### Runner3 (Windows)
 IP: 134.169.81.208
 Windows 11