Auszug |
---|
Visual Studio Code is is an IDE with full support for remote code development. |
...
Users install and run VS Code on their own machine. On the remote side no module is needed needs to be loaded - the client installs its own server automatically (~/.vscode-server
).
...
2) pick and fire up your remote compute node (avoid expansive tasks on logins), for example
Codeblock |
---|
salloc --partition=p cpu-clx |
3) add this block to your local SSH config to connect to any bcn*
compute node ( via blogin)
Codeblock | ||
---|---|---|
| ||
Host clx-login User <your_username> HostName blogin.nhr.zib.de IdentityFile ~/.ssh/<your_key> Host bcn* User <your_username> IdentityFile ~/.ssh/<your_key> ProxyCommand ssh -W %h:%p clx-login |
4) in VScode open the Command Palette, type/enter: Remote – SSH: Connect to Host, and provide the specific name of your allocated compute node, e.g. bcn####
As 5) as soon as the remote connection in VScode is established you can install additional extensions on the remote side, too. Some recommendations are
- GitHub Copilot
- Python
- JupyterHub
- C/C++
- Modern Fortran
Steps 2) and 4) need to be executed each time running VScode on a compute node. All other steps are required one time only - are permanent.
Optional: containerized VScode server
Advantage: inside container user has more rights, e.g., can use dnf
.