bug: self hosted VPC issues

[AndryHTC]

Provide environment information

-

Describe the bug

In addition to the magic link issue (#1384 (comment)), I'm also encountering connection errors while testing the self-hosted version, while it works with the cloud one.

npx trigger.dev@latest dev --env-file trigger/.env.local   

Trigger.dev (3.3.10)
------------------------------------------------------
Key: Version | Task | Run
------------------------------------------------------
○ Building background worker…
│
■  Error: Build failed: Connection error.
│
│         Error: Connection error.
│             at DevWorkerRuntime.initializeWorker (file:///C:/Users/Andry/AppData/Local/npm-cache/_npx/f51a09bd0abf5f10/node_modules/trigger.dev/dist/esm/dev/workerRuntime.js:153:19)
│             at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
│             at async updateBundle (file:///C:/Users/Andry/AppData/Local/npm-cache/_npx/f51a09bd0abf5f10/node_modules/trigger.dev/dist/esm/dev/devSession.js:49:13)
│             at async runBundle (file:///C:/Users/Andry/AppData/Local/npm-cache/_npx/f51a09bd0abf5f10/node_modules/trigger.dev/dist/esm/dev/devSession.js:119:9)
│             at async startDevSession (file:///C:/Users/Andry/AppData/Local/npm-cache/_npx/f51a09bd0abf5f10/node_modules/trigger.dev/dist/esm/dev/devSession.js:122:26)
│             at async startDev (file:///C:/Users/Andry/AppData/Local/npm-cache/_npx/f51a09bd0abf5f10/node_modules/trigger.dev/dist/esm/commands/dev.js:115:23)
│             at async devCommand (file:///C:/Users/Andry/AppData/Local/npm-cache/_npx/f51a09bd0abf5f10/node_modules/trigger.dev/dist/esm/commands/dev.js:55:29)
│             at async file:///C:/Users/Andry/AppData/Local/npm-cache/_npx/f51a09bd0abf5f10/node_modules/trigger.dev/dist/esm/commands/dev.js:31:13
│             at async file:///C:/Users/Andry/AppData/Local/npm-cache/_npx/f51a09bd0abf5f10/node_modules/trigger.dev/dist/esm/cli/common.js:54:28
│             at async wrapCommandAction (file:///C:/Users/Andry/AppData/Local/npm-cache/_npx/f51a09bd0abf5f10/node_modules/trigger.dev/dist/esm/cli/common.js:40:12)

With the debug level log enabled: the base URL for seems to be inconsistent, switching between api.trigger.dev, localhost:3040, and remote.server:3040.

npx trigger.dev@latest dev --env-file trigger/.env.local -l debug

apiUrl: 'https://api.trigger.dev',
// ...
login: {
  // ...
  dashboardUrl: 'http://localhost:3040',
// ...
Initializing dev environment for project proj_cviwyoihnaqwclotauwp http://remote.server:3040
// ...
initialized worker runtime {
  // ...
  client: CliApiClient {
    apiURL: 'http://localhost:3040',
  // ...

Have I missed something?
Could you provide some clarification on the expected base URL behavior within a VPC environment? Additionally, any insights into potential reasons for the connection errors would be helpful for troubleshooting.

Thanks again for your support!

Reproduction repo

-

To reproduce

1. Self host Trigger.dev in a remote VPC protected server
2. Login the CLI
3. Run the dev command

Additional information

No response

[AndryHTC]

In the README.md there is this:

If you want to self-host the Trigger.dev platform, when you're developing your web app locally you'll need to run the Trigger.dev platform locally as well.

Is it still valid?

Also:

You’re going to need at least one Debian (or derivative) machine with Docker and Docker Compose installed

Does it apply to the local one too?

What is the correct dev flow?