Add document about debugging offline state

[jsoref]

There's documentation for intentionally running Rancher Desktop offline, but there isn't documentation explaining how to find out why Rancher Desktop thinks it's offline when the user doesn't.

[Nino-K]

Currently, the built-in check in the electron app may not be robust enough to determine the network availability status and could yield into false positive scenarios.

More info can be found here and here in the documentation.

For users to determine the false positive "offline" state from a true "offline" e.g when running Rancher Desktop in an air-gapped scenario as described here they would need to perform a basic test. You can determine a true network availability by attempting to ping a well-known URL, e.g google.ca or in a browser navigate to a know working URL e.g release pages on GitHub or https://www.google.com/generate_204. If a successful response is received and Rancher Desktop is showing an offline state this is an indication of a false positive network state report from the built-in electron app (chrome).

[gunamata]

@jsoref , Would adding a FAQ as below help?

Why does Rancher Desktop show the network status as offline even when my machine is connected to the internet?

Rancher Desktop uses logic similar to browsers to determine if the machine is currently online. However, this can be influenced by several factors, such as a browser's internal connection management, firewall or proxy settings, other software installed on the machine etc. So, if Rancher Desktop shows the network status as offline, but if you can access the internet without issues, then please ignore the Network status message displayed in the General tab or elsewhere in the app.

[jsoref]

Maybe. I mean, it's definitely a start.

There are technically two reasons a user will care about this:

• The user's Rancher Desktop isn't working properly
  • they can't get kubernetes to install
  • they can't get kubernetes to start
  • they can't pull images
  • they can't make connections w/ nerdctl/docker
  • their programs can't access external resources once they start
• The user sees the idiot light and is curious

I'm generally in this last bullet. But that average user is more likely to be in the first bullet.

We regularly see people with VPNs, Firewalls, or Proxies that cause problems, and I'd assume that the indicator triggers for some of these.

Ideally your FAQ entry should give some information about how to debug the network to enable it to work, or a way to work around it.

There's current-ish problem where users can't install kubernetes 1.24 but can install 1.21 -- I don't remember the whys, but if the light triggers for that, you'd definitely want to mention it as well. And if the light doesn't trigger and there isn't a FAQ entry for this, then there needs to be one for it.

Does that make sense?