| Previous | Next |
| ENOPROTOOPT | EPROTO |
ENETDOWN
ENETDOWN is a local network-stack failure. It says that the interface needed to reach the destination is down, rather than that the remote service rejected a connection. A send or connect can therefore fail before a packet is delivered to the peer. This is especially relevant when an application has more than one route, runs in a network namespace, or uses a virtual, VPN, tunnel, or container interface.
Keep it separate from ENETUNREACH, which describes the absence of a route to a network, and from EHOSTUNREACH, which is normally more specific to a destination host. Bringing up an unrelated interface or restarting the remote service does not answer which egress interface and route the failing socket actually selected.
Evidence that identifies the failing path
- Record the operation, destination address, address family, local address if already bound, and the socket's network namespace.
- Run
ip route getfor the exact destination and inspect the chosen device, source address, route table, and policy rule. - Inspect that device with
ip link show; distinguish administrative state, carrier state, and the existence of the device in the namespace. - For a managed service, correlate the failure with interface, DHCP, VPN, container, or network-manager events rather than retrying the same request without a delay policy.
POSIX sendto() · Linux connect(2) · Linux ip-link(8)
Looking for a different code? Search another status or error code.
