Substrate-specific P2P networking.
Important: This crate is unstable and the API and usage may change.
In a decentralized network, each node possesses a network private key and a network public key. In Substrate, the keys are based on the ed25519 curve.
From a node’s public key, we can derive its identity. In Substrate and libp2p, a node’s
identity is represented with the
PeerId struct. All network communications between nodes on
the network use encryption derived from both sides’s keys, which means that identities cannot
A node’s identity uniquely identifies a machine on the network. If you start two or more clients using the same network key, large interferences will happen.
Substrate’s networking protocol is based upon libp2p. It is at the moment not possible and not planned to permit using something else than the libp2p network stack and the rust-libp2p library. However the libp2p framework is very flexible and the rust-libp2p library could be extended to support a wider range of protocols than what is offered by libp2p.
In order for our node to join a peer-to-peer network, it has to know a list of nodes that are part of said network. This includes nodes identities and their address (how to reach them). Building such a list is called the discovery mechanism. There are three mechanisms that Substrate uses:
- Bootstrap nodes. These are hard-coded node identities and addresses passed alongside with the network configuration.
- mDNS. We perform a UDP broadcast on the local network. Nodes that listen may respond with their identity. More info here. mDNS can be disabled in the network configuration.
- Kademlia random walk. Once connected, we perform random Kademlia
FIND_NODErequests on the configured Kademlia DHTs (one per configured chain protocol) in order for nodes to propagate to us their view of the network. More information about Kademlia can be found on Wikipedia.
When node Alice knows node Bob’s identity and address, it can establish a connection with Bob.
All connections must always use encryption and multiplexing. While some node addresses (eg.
/quic) already imply which encryption and/or multiplexing to use, for others
the multistream-select protocol is used in order to negotiate an encryption layer and/or a
The connection establishment mechanism is called the transport.
As of the writing of this documentation, the following base-layer protocols are supported by Substrate:
- TCP/IP for addresses of the form
/ip4/22.214.171.124/tcp/5. Once the TCP connection is open, an encryption and a multiplexing layer are negotiated on top.
- WebSockets for addresses of the form
/ip4/126.96.36.199/tcp/5/ws. A TCP/IP connection is open and the WebSockets protocol is negotiated on top. Communications then happen inside WebSockets data frames. Encryption and multiplexing are additionally negotiated again inside this channel.
- DNS for addresses of the form
/dns/example.com/tcp/5/ws. A node’s address can contain a domain name.
- (All of the above using IPv6 instead of IPv4.)
On top of the base-layer protocol, the Noise protocol is negotiated and applied. The exact handshake protocol is experimental and is subject to change.
The following multiplexing protocols are supported:
Once a connection has been established and uses multiplexing, substreams can be opened. When a substream is open, the multistream-select protocol is used to negotiate which protocol to use on that given substream.
Protocols that are specific to a certain chain have a
<protocol-id> in their name. This
“protocol ID” is defined in the chain specifications. For example, the protocol ID of Polkadot
is “dot”. In the protocol names below,
<protocol-id> must be replaced with the corresponding
Note: It is possible for the same connection to be used for multiple chains. For example, one can use both the
/sub/sync/2protocols on the same connection, provided that the remote supports them.
Substrate uses the following standard libp2p protocols:
/ipfs/ping/1.0.0. We periodically open an ephemeral substream in order to ping the remote and check whether the connection is still alive. Failure for the remote to reply leads to a disconnection.
/ipfs/id/1.0.0. We periodically open an ephemeral substream in order to ask information from the remote.
/<protocol_id>/kad. We periodically open ephemeral substreams for Kademlia random walk queries. Each Kademlia query is done in a separate substream.
Additionally, Substrate uses the following non-libp2p-standard protocols:
<protocol-id>must be replaced with the protocol ID of the targeted chain, and
<version>is a number between 2 and 6). For each connection we optionally keep an additional substream for all Substrate-based communications alive. This protocol is considered legacy, and is progressively being replaced with alternatives. This is designated as “The legacy Substrate substream” in this documentation. See below for more details.
/<protocol-id>/sync/2is a request-response protocol (see below) that lets one perform requests for information about blocks. Each request is the encoding of a
BlockRequestand each response is the encoding of a
BlockResponse, as defined in the
api.v1.protofile in this source tree.
/<protocol-id>/light/2is a request-response protocol (see below) that lets one perform light-client-related requests for information about the state. Each request is the encoding of a
light::Requestand each response is the encoding of a
light::Response, as defined in the
light.v1.protofile in this source tree.
/<protocol-id>/transactions/1is a notifications protocol (see below) where transactions are pushed to other nodes. The handshake is empty on both sides. The message format is a SCALE-encoded list of transactions, where each transaction is an opaque list of bytes.
/<protocol-id>/block-announces/1is a notifications protocol (see below) where block announces are pushed to other nodes. The handshake is empty on both sides. The message format is a SCALE-encoded tuple containing a block header followed with an opaque list of bytes containing some data associated with this block announcement, e.g. a candidate message.
- Notifications protocols that are registered using
NetworkConfiguration::notifications_protocols. For example:
/paritytech/grandpa/1. See below for more information.
Substrate uses a component named the peerset manager (PSM). Through the discovery mechanism, the PSM is aware of the nodes that are part of the network and decides which nodes we should perform Substrate-based communications with. For these nodes, we open a connection if necessary and open a unique substream for Substrate-based communications. If the PSM decides that we should disconnect a node, then that substream is closed.
For more information about the PSM, see the sc-peerset crate.
Note that at the moment there is no mechanism in place to solve the issues that arise where the two sides of a connection open the unique substream simultaneously. In order to not run into issues, only the dialer of a connection is allowed to open the unique substream. When the substream is closed, the entire connection is closed as well. This is a bug that will be resolved by deprecating the protocol entirely.
Within the unique Substrate substream, messages encoded using
parity-scale-codec are exchanged.
The detail of theses messages is not totally in place, but they can be found in the
Once the substream is open, the first step is an exchange of a status message from both sides, containing information such as the chain root hash, head of chain, and so on.
Communications within this substream include:
- Syncing. Blocks are announced and requested from other nodes.
- Light-client requests. When a light client requires information, a random node we have a substream open with is chosen, and the information is requested from it.
- Gossiping. Used for example by grandpa.
A so-called request-response protocol is defined as follow:
- When a substream is opened, the opening side sends a message whose content is protocol-specific. The message must be prefixed with an LEB128-encoded number indicating its length. After the message has been sent, the writing side is closed.
- The remote sends back the response prefixed with a LEB128-encoded length, and closes its side as well.
Each request is performed in a new separate substream.
A so-called notifications protocol is defined as follow:
- When a substream is opened, the opening side sends a handshake message whose content is
protocol-specific. The handshake message must be prefixed with an
LEB128-encoded number indicating its length. The
handshake message can be of length 0, in which case the sender has to send a single
- The receiver then either immediately closes the substream, or answers with its own
LEB128-prefixed protocol-specific handshake response. The message can be of length 0, in which
case a single
0has to be sent back.
- Once the handshake has completed, the notifications protocol is unidirectional. Only the node which initiated the substream can push notifications. If the remote wants to send notifications as well, it has to open its own undirectional substream.
- Each notification must be prefixed with an LEB128-encoded length. The encoding of the messages is specific to each protocol.
- Either party can signal that it doesn’t want a notifications substream anymore by closing its writing side. The other party should respond by closing its own writing side soon after.
The API of
sc-network allows one to register user-defined notification protocols.
sc-network automatically tries to open a substream towards each node for which the legacy
Substream substream is open. The handshake is then performed automatically.
For example, the
sc-finality-grandpa crate registers the
At the moment, for backwards-compatibility, notification protocols are tied to the legacy Substrate substream. Additionally, the handshake message is hardcoded to be a single 8-bits integer representing the role of the node:
- 1 for a full node.
- 2 for a light node.
- 4 for an authority.
In the future, though, these restrictions will be removed.
sc-network crate is done through the
NetworkWorker struct. Create this
struct by passing a
config::Params, then poll it as if it was a
Future. You can extract an
Arc<NetworkService> from the
NetworkWorker, which can be shared amongst multiple places
in order to give orders to the networking.
config module for more information about how to configure the networking.
NetworkWorker has been created, the important things to do are:
NetworkWorker::pollin order to advance the network. This can be done by dispatching a background task with the
on_block_importwhenever a block is added to the client.
on_block_finalizedwhenever a block is finalized.
trigger_repropagatewhen a transaction is added to the pool.
More precise usage details are still being worked on and will likely change in the future.
Bitswap server for substrate.
Helper for handling (i.e. answering) block requests from a remote peer via the
Configuration of the networking layer.
Substrate network possible errors.
Helper for sending rate-limited gossip messages.
Helpers for outgoing and incoming light client requests.
Information about the networking, for diagnostic purposes.
Helper for handling (i.e. answering) state requests from a remote peer via the
Transactions handling to plug on top of the network service.
Representation of a Multiaddr.
Substrate network service. Handles network IO and manages connectivity.
Overview status of the network.
Main network worker. Must be polled in order for the network to advance.
Reserved slot in the notifications buffer, ready to accept data.
Identifier of a peer of the network.
Info about a peer’s known state.
Description of a reputation adjustment for a node.
Reported state download progress.
Events generated by DHT as a response to get_value and put_value requests.
Type for events generated by networking layer.
When sending a request, what to do on a disconnected recipient.
Role that the peer sent to us during the handshake, with the addition of what our local node knows about that peer.
Possible failures occurring in the context of sending an outbound request and receiving the response.
Error in a request.
Reported sync state.
Minimum Requirements for a Hash within Networking
Trait for providing information about the local network state