Running Swarm for the first time

1. Check the swarm command is in the PATH

Verify this with which swarm.

If the swarm command cannot be found, you can add the Swarm binary route to your PATH for easier access.

Alternatively, you will need to use $GOPATH/bin/swarm instead, assuming you are compiling directly from the source code.

2. Execute the swarm command

To boot up Swarm for the first time simply execute:


3. Specify a private key

You will need a private key (sometimes referred to as account) to use Swarm. This can either:

  • be specified by the bzzaccount flag (if the account already exists)
  • be created through a Swarm prompt (if not specified)

Note that after used once, an account will continue to be used by default if not specified the next time.

4. Verify the client is running

If Swarm was started successfully, the local web server endpoint should be accessible through your browser, by default at http://localhost:8500.

This web interface will allow you to upload and download files manually.

In contrast, the terminal in which swarm was executed will allow you to take a look at log outputs and relevant execution messages, as well as errors.

5. CLI interaction

The terminal can be used to interact with Swarm once the system is up and running. The 2 most basic commands are:

  • up
    • e.g. swarm up dog.png
    • the output will look something like 858ce596aedec06c4d31d983dcc8a0df404b3c925600a8a9ac0332cc296b7883. This is (in essence) the content hash for the uploaded file, which can then be used to retrieve it.
  • down
    • e.g. swarm down bzz://858ce596aedec06c4d31d983dcc8a0df404b3c925600a8a9ac0332cc296b7883
    • if successful, there will be no output, and the corresponding file according to this hash will be placed in the current directory.

You can find a list of commands here.

Receive updates

Get the latest updates from the Rootstock ecosystem