python/cpython · error · ValueError

ssl_shutdown_timeout is only meaningful with ssl

Error message

ssl_shutdown_timeout is only meaningful with ssl

What it means

Raised by create_connection when ssl_shutdown_timeout is given but ssl is not. This timeout bounds the TLS shutdown (close_notify) phase of the connection; like the handshake timeout it is meaningful only for TLS, so asyncio rejects it on plaintext connections.

Source

Thrown at Lib/asyncio/base_events.py:1110

            # already-connected socket was passed or when only a port
            # is given.  To avoid this error, you can pass
            # server_hostname='' -- this will bypass the hostname
            # check.  (This also means that if host is a numeric
            # IP/IPv6 address, we will attempt to verify that exact
            # address; this will probably fail, but it is possible to
            # create a certificate for a specific IP address, so we
            # don't judge it here.)
            if not host:
                raise ValueError('You must set server_hostname '
                                 'when using ssl without a host')
            server_hostname = host

        if ssl_handshake_timeout is not None and not ssl:
            raise ValueError(
                'ssl_handshake_timeout is only meaningful with ssl')

        if ssl_shutdown_timeout is not None and not ssl:
            raise ValueError(
                'ssl_shutdown_timeout is only meaningful with ssl')

        if sock is not None:
            _check_ssl_socket(sock)

        if happy_eyeballs_delay is not None and interleave is None:
            # If using happy eyeballs, default to interleave addresses by family
            interleave = 1

        if host is not None or port is not None:
            if sock is not None:
                raise ValueError(
                    'host/port and sock can not be specified at the same time')

            infos = await self._ensure_resolved(
                (host, port), family=family,
                type=socket.SOCK_STREAM, proto=proto, flags=flags, loop=self)
            if not infos:

View on GitHub (pinned to bc6749cc3b)

Solutions

  1. Pass ssl_shutdown_timeout only in the branch that also passes ssl=ctx.
  2. Strip TLS-only keys from kwargs when TLS is disabled: kwargs.pop('ssl_shutdown_timeout', None).
  3. Keep ssl, server_hostname, and both ssl timeouts as one atomic kwargs group.

Example fix

// before
await loop.create_connection(proto, h, p, ssl_shutdown_timeout=5)  # no ssl

// after
await loop.create_connection(proto, h, p, ssl=ctx, ssl_shutdown_timeout=5)
Defensive patterns

Strategy: validation

Validate before calling

tls_kwargs = {}
if ctx is not None:
    tls_kwargs.update(ssl=ctx, ssl_shutdown_timeout=5)

Try / catch

try:
    await loop.create_connection(proto, h, p, ssl_shutdown_timeout=5)
except ValueError:
    await loop.create_connection(proto, h, p, ssl=ssl.create_default_context(),
                                 ssl_shutdown_timeout=5)

Prevention

When it happens

Trigger: Calling loop.create_connection(proto, host, port, ssl_shutdown_timeout=5.0) with ssl unset (None). The argument exists only in newer Pythons; combined with conditional TLS it is easy to leak into a plaintext call.

Common situations: Config-driven client builders that forward a TLS settings object as **kwargs and a code path that sets ssl=None; disabling TLS against a local test server while keeping the shutdown timeout setting.

Understand the failure class

Related errors


AI-assisted analysis of python/cpython@bc6749cc3b (2026-08-14). Data as JSON: /api/errors/17e6dc9e8359dcce. Report an issue: GitHub.