Server API

aiosignalr.server

SignalRServer

The hub server. See the server guide for usage.

SignalRServer(
    hub: type[Hub],
    *,
    options: ServerOptions | None = None,
    logger: logging.Logger | None = None,
)

Methods & attributes

Member	Description
protocol_registry	{"json": JSONProtocol, "messagepack": MessagePackProtocol}
lifetime	HubLifetimeManager used for broadcasts/groups
dispatcher	HubDispatcher that binds hub methods
async serve(host, port, *, path="/hub") -> tuple[str, int]	Run standalone aiohttp server
create_application(*, path="/hub")	Build the aiohttp app
asgi_app(*, path="/hub")	Build an ASGI app for uvicorn/FastAPI
async close()	Shut down the standalone server
get_connection(connection_id)	Look up a live connection context

ServerOptions

Field	Default	Description
keep_alive_interval	15.0	Server ping interval
client_timeout_interval	30.0	Silent-client timeout
handshake_timeout	15.0	Handshake deadline
maximum_parallel_invocations	1	Concurrent invocations per connection
maximum_receive_message_size	32768	Max inbound frame size
stream_buffer_capacity	10	Upload-stream queue capacity
long_polling_timeout	15.0	Long poll duration
enable_detailed_errors	False	Send exception details to clients
allow_stateful_reconnects	False	Enable stateful reconnect
stateful_reconnect_buffer_size	100_000	Outbound buffer bytes
stateful_reconnect_timeout	30.0	Seconds a dropped connection is kept

Hub

Base class for hubs. Public methods become invocable; async generators become streaming methods.

• self.clients — HubClients proxy (see below).
• self.groups — GroupsManager for the caller's membership.
• self.context — HubContext with connection_id, headers, query, user_id.
• async on_connected() — per-connection hook.
• async on_disconnected(exception) — per-connection hook.

HubClients

Member	Description
all_	Proxy for every client (send(method, *args))
others	Every client except the caller
connection(id)	A single connection
connections(ids)	Several connections
group(name)	Members of a group
groups(names)	Members of several groups
user(id)	Connections with a user id
users(ids)	Connections with several user ids
caller	The calling client (invoke(...) supports client results)

HubLifetimeManager / DefaultHubLifetimeManager

Abstract registry used for broadcasts and groups. DefaultHubLifetimeManager is the in-memory implementation: it tracks connections, maintains groups and serializes each broadcast once per protocol. Subclass it to plug in a distributed backplane.

HubConnectionContext

Per-connection server state (connection id, negotiated protocol, outbound transport, message buffer, invocation tracking). Normally you interact with it only via self.context in hub methods or custom lifetime/dispatcher code.

• has_transport — whether an outbound transport is attached.
• connection_id — logical connection id.
• use_stateful_reconnect — whether the ack protocol is active.
• async send(message) / send_bytes(data) / send_serialized(data) — outbound writes (the buffer-aware path).
• wait_closed() — resolves when the connection closes.

IncomingStream

Async-iterable bound to an upload-stream parameter (async for item in stream).

StreamTracker

Tracks client-to-server upload streams for a connection.

HubDispatcher

Discovers hub methods and routes messages. See method resolution rules.

asgi_app

asgi_app(server, *, path="/hub") — returns a plain ASGI app; also available as server.asgi_app().