I regularly find myself running a python script to update my ~/.ssh/config entries, whenever I create new servers on the Faculty platform. I thought it might be useful to contribute back and add it as a feature to the CLI.

Problem statment

Currently, if you create a new server and would like to access it remotely via SSH, you have to run a few separate commands:

1️⃣ Retrieve a list of your servers by name and (only with the verbose flag) their Project:

> faculty server list -v                                   
Project Name        Server Name       Type     Machine Type      CPUs  RAM    Status    Server ID            Started
My Project Name     server_name       jupyter  -                    1  4GB    running   some-server-id       2023-01-01 01:23

2️⃣ Retrieve an IP address for my server:

> faculty server ssh-details "Project Name" server_name
Hostname       Port  Username
12.34.56.78   12345  username

3️⃣ Copy-paste that IP address from my std output and then run my ssh command

> ssh [email protected] -P 12345

This is a little cumbersome for workflows where you create/update your servers often.

Solution

My proposed change is to modify the faculty server list command, adding an extra --format option which supports returning a list of servers in a format which can be piped directly to a user's ssh_config file.
This is super useful for workflows where you create/update your servers often, since it allows you to "save" your server ssh details and propagate that info to places like your IDE or command line client.

Example:

> faculty server list  --format ssh-config
Host server_name-project_name
    HostName 12.34.56.78
    User username
    Port 12345

I've removed the --verbose flag, and moved retrieving the full table of server info to --format table, since it seemed overcomplicated to have both "verbose" and "format" tags.