Creating Streams

See Authentication with the CLI page on how to authenticate.

Creating a stream

A stream can be created as follows:

$ strm create stream demo -o json
{
  "ref": {
    "name": "demo",
    "projectId": "30fcd008-9696-...."
  },
  "enabled": true,
  "limits": {
    "eventRate": "99",
    "eventCount": "999999"
  },
  "credentials": [
    {
      "clientId": "stream-ylbt4v9o6dvvc...",
      "clientSecret": "M0fBiQnKNXn*U...",
    }
  ],
  "maskedFields": {}
}

The clientId and clientSecret pair is what identifies your stream when you send data to STRM Privacy. STRM Privacy uses the OAuth 2.0 client credentials flow to generate a bearer token that needs to be provided with each HTTP request. Our drivers have tooling to create and refresh these tokens, but nothing prevents you from creating the headers by hand (see the page about sending data by hand to do this manually).

tip

Use the --save flag to store the credentials on your local machine in ~/.config/strmprivacy/saved-entities/Stream/(stream-name).json

So with this you have enough information to start sending data to events.strmprivacy.io/event. With the same credentials you can connect to the web-socket endpoint with a websocket client to receive the events as you send them. Use strm listen web-socket <stream-name> [--client-id .. --client-secret ..]. See here for details.

Listing streams

Streams can be listed and will be shown as a table.

$ strm list streams
 STREAM   DERIVED   PURPOSES   ENABLED

 demo     false     []         true

To show more info, use the --output flag, and try out different formats, such as json.

$ strm list streams --output json
{
    "streams": [
        {
            "stream": {
                "ref": {
                    "name": "demo",
                    "projectId": "30fcd008-9696-...."
                },
                "enabled": true,
                "limits": {
                    "eventRate": "99",
                    "eventCount": "999999"
                },
                "credentials": [
                    {
                        "clientId": "stream-ylbt4v9o6dvvc...",
                        "clientSecret": "M0fBiQnKNXn*U...",
                    }
                ],
                "maskedFields": {
                  "seed": "***"
                }
            }
        }
    ]
}

Creating decrypted streams

If you want STRM Privacy to decrypt data for certain data purposes, you need to create a privacy stream derived from an existing encrypted stream.

So let’s create one, in this case for the purposes 0 and 1 (refer to your purpose map to know which values to use):

$ strm create stream --derived-from demo --purposes 0,1
 STREAM     DERIVED   PURPOSES   ENABLED 
   
 demo-0-1   true      [0 1]      true                 

The derived stream is provided with a default name demo-0-1 because we did not provide an explicit name. If we had added a name after the strm create stream we would have created a stream with that name.

The derived stream demo-0-1 captures data from the encrypted source stream demo. It will drop all events that have not been allowed to be used for the configured purposes (for example due to a data subject not consenting). The events that do have permission to be processed, will be (partially) decrypted: fields filed under these purposes will be decrypted, while any other sensitive fields will remain encrypted.

Cleaning up

Delete a stream, all its dependents and all its data (that hasn’t been exported yet) with the --recursive flag:

$ strm delete stream demo --recursive
Stream has been deleted