docs: completed v1 docs (#36)

This commit is contained in:
Morten Olsen
2024-01-15 15:10:19 +01:00
committed by GitHub
parent 161a098c9f
commit 1c3b993ab2
27 changed files with 226 additions and 41 deletions

View File

@@ -0,0 +1,49 @@
## Getting Started with mini loader CLI
Welcome to the mini loader CLI! This guide will walk you through the installation of the CLI, creating a simple script, and executing it locally to ensure everything works smoothly.
### Step 1: Install the CLI
The mini loader CLI is your gateway to interacting with a mini-loader server and running workloads locally for validation. Install it globally using npm with the following command:
```bash
npm install -g @morten-olsen/mini-loader-cli
```
### Step 2: Create Your First Script
Now, let's write a basic script that outputs a single artifact named “hello”. Create a new file with the following JavaScript code:
```javascript
import { artifacts } from '@morten-olsen/mini-loader';
const run = async () => {
artifacts.create('hello', 'world');
};
run();
```
Save this file as `script.js`.
#### A Note on Dependencies
In this script, we're using the `@morten-olsen/mini-loader` package, which might not be installed in your local environment. No worries though, as mini loader can automatically download necessary packages when preparing the script. Alternatively, for a more structured approach (especially if you're using TypeScript), you can initialize a Node.js project and install the dependencies for complete access to typings.
### Step 3: Run the Script Locally
To validate that your script is functioning correctly, execute it locally using the following command:
```bash
mini-loader local run script.js -ai
```
The `-ai` flag instructs the CLI to automatically download any referenced packages when bundling the script.
After running the command, you should see an output confirming that a new artifact named “hello” was created successfully.
### What's Next?
Congratulations on setting up and running your first script with mini loader! You're now ready to take the next step.
[Next: Setting Up the Server](./02-setup-server.md)

View File

@@ -0,0 +1,61 @@
Certainly! Here's a revised version of your documentation page to make it
## Quick Start with mini loader using Docker
This guide will help you quickly set up and run a mini loader server using Docker. Follow these simple steps to deploy your server and start interacting with it using the [mini-loader CLI](./01-first-workload.md).
### Step 1: Deploy the mini loader Container
To begin, let's deploy the mini loader container. Run the following command in your terminal:
```bash
docker run -p 4500:4500 -n mini-loader ghcr.io/morten-olsen/mini-loader:latest
```
This command downloads the latest mini loader image and runs it, exposing port 4500.
### Step 2: Verify Server Health
Once the container is running, you can check the server's health:
```bash
curl http://localhost:4500/health
```
This command should return a response indicating that the server is running smoothly.
### Step 3: Authorize Using a Token
Initially, your server won't have any authorization setup. You can issue a token directly from the container to use for authorization:
```bash
docker exec mini-loader mini-loader create-token
```
This command will output a token. Keep this token handy as you'll need it for the next step.
### Step 4: Authorize the CLI Client
Now, authorize your CLI client to interact with the server:
```bash
mini-loader auth login https://localhost:4500
```
Enter the token when prompted.
### Step 5: Verify CLI Connection
Finally, verify that the CLI is properly configured to interact with your server:
```bash
mini-loader loads ls
```
This command lists all the loads currently on your server, confirming that the CLI is communicating successfully with the server.
### Ready to Go!
You've successfully deployed and configured your mini loader server using Docker! You're now ready to start interacting with the server.
[Next: Interacting with the Server](./03-interacting-with-server.md)

View File

@@ -0,0 +1,70 @@
## Interacting with the server
The mini loader CLI is your primary tool for interacting with the server. This guide will walk you through the basic commands and their usage.
### Getting Help
For assistance with any command, append `--help` to it. For example, to see documentation for the `loads push` command:
```bash
mini-loader loads push --help
```
### Push a Load to the Server
To push a script to your server:
```bash
mini-loader loads push <script-path> -i <id>
```
- `<id>` is the identifier for your script. If omitted, an ID will be auto-generated.
- To immediately trigger a run of the load after pushing, add the `-r` flag.
### Checking Run Status
To list the status of all runs:
```bash
mini-loader runs ls
```
- To filter runs by a specific load, add `-l <load-id>`.
### Retrieving Logs
To view logs:
```bash
mini-loader logs ls
```
- To view logs from a specific run, use `-r <run-id>`.
- To view logs for a specific load, use `-l <load-id>`.
### Listing Artifacts
To list all artifacts:
```bash
mini-loader artifacts ls
```
- You can filter the artifacts by a specific run using `-r <run-id>`.
- Alternatively, filter by load using `-l <load-id>`.
### Downloading an Artifact
To download a specific artifact:
```bash
mini-loader artifacts pull <id> myfile.txt
```
Replace `<id>` with the identifier of the artifact you wish to download.
### Ready for More?
You're now equipped to manage loads, runs, logs, and artifacts using the mini loader CLI. For advanced usage, such as managing secrets, proceed to the next section.
[Next: Managing Secrets](./04-managing-secrets.md)

View File

@@ -0,0 +1,81 @@
## Managing Secrets
### Introduction
In many workflows, accessing sensitive data such as API tokens or credentials is essential. To handle this securely, you can use secrets management. This tutorial demonstrates how to manage secrets using the CLI and implement them in a simple Node.js workload.
### Creating Secrets with the CLI
To create a new secret, use the `mini-loader` CLI as follows:
```bash
mini-loader secrets set <id>
```
For example, to store a GitHub personal access token, you would use:
```bash
mini-loader secrets set githubtoken
```
Upon execution, you'll be prompted to enter your access token.
### Implementing Secrets in Your Workload
Next, let's create a Node.js script (`github.js`) that uses this token to fetch your GitHub username and saves it as an artifact.
1. **Create `github.js` File:**
```javascript
import { secrets, artifacts } from '@morten-olsen/mini-loader';
import { Octokit } from '@octokit/rest';
// Retrieve the secret
const accessToken = secrets.get('githubtoken');
// Main async function to fetch and save GitHub username
async function run() {
const octokit = new Octokit({ auth: accessToken });
const user = await octokit.users.getAuthenticated();
await artifacts.create('user', JSON.stringify(user.data.login));
}
// Execute the function
run().catch(console.error);
```
This script initializes the Octokit client with the access token, fetches the authenticated user's data, and then saves the username as an artifact.
2. **Run the Script:**
Execute your script with `mini-loader`:
```bash
mini-loader loads push github.js -r -ai
```
### Managing Local Secrets
If you're running the script locally, you can manage secrets either by using a `.secrets` file or setting an environment variable.
1. **Using a `.secrets` File:**
Create a file named `.secrets` and add your token:
```
githubtoken=<your-token>
```
2. **Using Environment Variables:**
Prefix your environment variable with `ML_S_` and run the script:
```bash
ML_S_githubtoken=<your-token> mini-loader local run github.js -ai
```
### Conclusion
By following these steps, you can securely manage and use secrets within your workloads, enhancing the security and integrity of your applications.
[Next: Creating an API](./05-creating-an-api.md)

View File

@@ -0,0 +1,52 @@
## Creating an API Inside Your Workload
Workloads in mini loader can set up simple HTTP servers by connecting to a socket file, a feature supported by many JavaScript server libraries.
### Binding Your Workload to an HTTP Endpoint
To expose your workload as an HTTP server, specify the path parameter using the `getPath()` method provided by the `@morten-olsen/mini-loader` package. This method dynamically assigns a path for your API.
### Important Note
Please be aware that the gateway provided by mini loader isn't fully featured. As such, certain functionalities like streaming and WebSockets may not be supported.
### Example: Setting Up a Server with Fastify
Here's how you can create a simple API server using Fastify in TypeScript:
```typescript
import { http } from '@morten-olsen/mini-loader';
import fastify from 'fastify';
const server = fastify();
// Handling all requests and returning the requested URL
server.all('*', async (req) => {
return req.url;
});
// Listening on the path provided by mini loader
server.listen({
path: http.getPath(),
});
```
With this setup, your server will respond to all incoming requests by returning the requested URL.
### Deploying Your Workload
Now, you can push and run your workload just like any other script:
```bash
mini-loader loads push -r my-script.ts
```
### Accessing Your Server
After pushing your workload, mini loader will display the run ID. You can use this ID to access your server. For example, to make a request to your server, you can use `curl`:
```bash
curl http://localhost:4500/gateway/{your-run-id}
```
Replace `{your-run-id}` with the actual run ID provided by mini loader.