README.md

Tideflow's Self-hosted Runner

With Tideflow's Self-hosted Runner, you can run workflow's task in your own server via TideFlow

For more information, visit TideFlow.io

Usage

npx @tideflowio/tideflow-agent -t [token] -u tideflow.example.com

$ tideflow-agent --help
Usage: index [options]

Options:
  -v, --version                       output the version number
  -c, --concurrency [concurrency]     Max number of jobs the agent should process concurrently
  -t, --token [token]                 Authentication token
  -u, --url [url]                     Tideflow url
  --noupdate                          Opt-out of update version check
  -h, --help                          output usage information

Examples:
  $ tideflow-agent --help
  $ tideflow-agent -h

Environment variables

# Specify authentication token.
# Optional. Having the authentication token stored as an environment
# variable allows users to run the agent without passing the -t parameter.
# Example: d2a04f78-ff8a-4eb4-a12c-57fb7abf03a7
TIDEFLOW_AGENT_TOKEN

# Specify the URL to connect to the Tideflow's platform.
# Optional. Having the URL stored as an environment
# variable allows users to run the agent without passing the -u parameter.
# Example: http://subdomain.example.com:3000
TIDEFLOW_AGENT_URL

How to process data from previous tasks

• type: an string that defines the kind of data retuned by the previous step (object, array, file, etc)
• data: an object containing the task's result.

Command actions

It's possible that the agent will receive preivous tasks results. This is defined in the workflow's task parameters.

The result from previous tasks are stored in a temporal file. This file's absolute path is passed as a parameter called "tideflow_previous_file" to the command to be executed.

For example, if the command to run is meow, the agent will execute it as:

meow --tideflow_previous_file "/tmp/random_file_name"

The content of such file is a JSON object, with a format similar to:

{
  "execution": {
    "_id": "neL9r33vm38txn9e6"
  },
  "tasks": {
    "the-task-id": {
      "_id": "iesFwrdTtBToRRBzB",
      "stepIndex": 1,
      "type": "file",
      "event": "read-file",
      "createdAt": "2020-10-03T15:55:08.637Z",
      "status": "success",
      "result": {
        "files": [
          {
            "fileName": "my-first-workflow-ifg.json",
            "url": "http://localhost:3000/download?type=actionFile&token=eyJhbGciO2A"
          }
        ]
      },
      "updatedAt": "2020-10-03T15:55:08.650Z"
    },
    "trigger": {
      "_id": "qJucoXaWaEJqcazoe",
      "stepIndex": "trigger",
      "type": "endpoint",
      "event": "called",
      "createdAt": "2020-10-03T15:55:08.629Z",
      "status": "success",
      "result": {
        "data": {
          "name": "Jose",
          "location": {
            "lat": 42.8867647,
            "lon": -9.2716399
          }
        }
      },
      "updatedAt": "2020-10-03T15:55:08.633Z"
    }
  }
}

NodeJS SFC actions

The result from previous tasks is passed as the handler's first argument.

module.exports.handler = previousResults => {
  // My logic here
  return { greeting: 'hello world' }
}

---

Contributing

If you would like to contribute to Tideflow, check out the Contributing Guide.

License

GNU GENERAL PUBLIC LICENSE

Developer Resources

• Documentation: https://docs.tideflow.io
• Contribute: https://docs.tideflow.io/docs/contribute