Scroll to navigation

IP-VRF(8) Linux IP-VRF(8)


ip-vrf - run a command against a vrf


ip vrf { COMMAND | help }

ip vrf show [ NAME ]

ip vrf identify [ PID ]

ip vrf pids NAME

ip vrf exec [ NAME ] command...


A VRF provides traffic isolation at layer 3 for routing, similar to how a VLAN is used to isolate traffic at layer 2. Fundamentally, a VRF is a separate routing table. Network devices are associated with a VRF by enslaving the device to the VRF. At that point network addresses assigned to the device are local to the VRF with host and connected routes moved to the table associated with the VRF.

A process can specify a VRF using several APIs -- binding the socket to the VRF device using SO_BINDTODEVICE, setting the VRF association using IP_UNICAST_IF or IPV6_UNICAST_IF, or specifying the VRF for a specific message using IP_PKTINFO or IPV6_PKTINFO.

By default a process is not bound to any VRF. An association can be set explicitly by making the program use one of the APIs mentioned above or implicitly using a helper to set SO_BINDTODEVICE for all IPv4 and IPv6 sockets (AF_INET and AF_INET6) when the socket is created. This ip-vrf command is a helper to run a command against a specific VRF with the VRF association inherited parent to child.

This command lists all VRF and their corresponding table ids. If NAME is given, then only that VRF and table id is shown. The latter command is useful for scripting where the table id for a VRF is needed.

This command allows applications that are VRF unaware to be run against a VRF other than the default VRF (main table). A command can be run against the default VRF by passing the "default" as the VRF name. This is useful if the current shell is associated with another VRF (e.g, Management VRF).

This command requires the system to be booted with cgroup v2 (e.g. with systemd, add systemd.unified_cgroup_hierarchy=1 to the kernel command line).

This command also requires to be run as root. Alternatively it can be run by an unprivileged user if the following capabilities(7) are given:

To load the BPF program.
To set the socket into the cgroup.
To create the cgroup subdir in /sys.

If these capabilities are added and if ip(8) is built with libcap(3) then these capabilities will be dropped before cmd is executed by ip vrf exec. For every other unprivileged invocation of ip(8) all capabilities will be dropped.

NOTE: capabilities will NOT be dropped if CAP_NET_ADMIN is set to INHERITABLE to avoid breaking programs with ambient capabilities that call ip.

This command shows the VRF association of the specified process. If PID is not specified then the id of the current process is used.

This command shows all process ids that are associated with the given VRF.


This command requires a kernel compiled with CGROUPS and CGROUP_BPF enabled.

The VRF helper *only* affects network layer sockets.


ip vrf exec red ssh

Executes ssh to against the VRF red table.


ip(8), ip-link(8), ip-address(8), ip-route(8), ip-neighbor(8)


Original Manpage by David Ahern

7 Dec 2016 iproute2