OpenBSD manual page server

Manual Page Search Parameters

VEB(4) Device Drivers Manual VEB(4)

veb, vportVirtual Ethernet Bridge network device

pseudo-device veb

The veb pseudo-device supports the creation of a single layer 2 Ethernet network between multiple ports. Ethernet interfaces are added to the veb bridge to be used as ports. Unlike bridge(4), veb takes over the operation of the interfaces that are added as ports. They are then independent of the host network stack: the individual Ethernet ports no longer function as independent devices and cannot be configured with inet(4) or inet6(4) addresses or other layer-3 features themselves.

The Ethernet network managed by veb can be connected to the network stack by creating a vport interface and adding it as a port to the bridge. From the perspective of the host network stack, a vport interface acts as a normal interface connected to an Ethernet network and can be configured with addresses.

veb is an IEEE 802.1Q VLAN aware bridge. A single veb instance can manage multiple independent and isolated Ethernet networks scoped by VLAN identifiers. Every packet traversing the bridge is associated with a VLAN, either by the identifier in 802.1Q VLAN tagged packets or by the VLAN identifier assigned to port interfaces for untagged packets.

Multiple vport interfaces can be attached to a single bridge to connect the network stack to different VLANs within the bridge, or to connect multiple routing domains to a single VLAN.

veb is a learning bridge that maintains a database of Ethernet addresses and the port that each address in a VLAN is reachable with. The bridge learns about the reachability of Ethernet addresses by reading the source address on packets received by ports, and then entering the address and port into the table dynamically. Static address entries may also be configured in the table, disabling dynamic learning for that address. Ethernet address learning can be disabled on individual ports.

When forwarding a packet, the address table is searched for the destination Ethernet address in the relevant VLAN and the packet is sent to the associated port in the table entry. If no entry is found in the table, or the packet is addressed to a multicast or broadcast address, the packet is flooded to all other ports on the bridge. Flooding of packets to unknown unicast addresses can be disabled on individual ports.

veb provides multiple mechanisms for filtering packets traversing the bridge.

By default veb filters IEEE 802.1Q VLAN tagged packets on a port until a set of allowed VLAN identifiers is configured. Packets without IEEE 802.1Q VLAN tags are assigned the native VLAN identifier configured on the port for untagged packets. Alternatively, a port can be configured to drop untagged packets, or have processing declined by the bridge and passed through to the network stack.

Ports can be configured as members of protected domains to restrict communication between them. This can be used to construct Virtual Private LAN Service (VPLS) topologies with protected domains enforcing a split-horizon forwarding rule. veb supports RFC 5517 Private VLANs to provide isolation between ports within a VLAN, and as part of a larger PVLAN network topology. The PVLAN implementation also provides support for configuring ports with the pvptags flag to communicate with VLAN aware but PVLAN unaware equipment using only primary VLAN identifiers for VLAN tagged packets.

veb can filter Ethernet packets entering or leaving ports using bridge rules. Ports can be configured to only allow relaying of IP (and ARP/RARP) packets by setting the blocknonip flag.

pf(4) can be used to filter IP packets as they enter or leave the bridge. By default this filtering is disabled, but can be enabled by setting the link1 flag. This filtering only applies to untagged packets on ports. The exception to this policy is on vport interfaces, where pf(4) runs as packets enter and leave the network stack regardless of the value of the link1 flag. A consequence of this behaviour is that packets traversing vport interfaces appear to travel in the opposite direction to packets travelling over other ports.

veb supports the addition of span ports to the bridge. Span ports transmit a copy of every packet received by the bridge, allowing for passive monitoring of traffic on a separate host.

ipsec(4), options(4), pf(4), hostname.if(5), ifconfig(8), netstart(8)

IEEE 802.1Q standard, https://standards.ieee.org/standard/802_1Q-2018.html.

S. HomChaudhuri and M. Foschiano, Cisco Systems' Private VLANs: Scalable Security in a Multi-Client Environment, RFC 5517, February 2010.

The veb driver first appeared in OpenBSD 6.9. VLAN and PVLAN support was added in OpenBSD 7.9.

David Gwynne <dlg@openbsd.org>

November 25, 2025 OpenBSD-current