VRF¶
VRF devices combined with ip rules provides the ability to create virtual routing and forwarding domains (aka VRFs, VRF-lite to be specific) in the Linux network stack. One use case is the multi-tenancy problem where each tenant has their own unique routing tables and in the very least need different default gateways.
Configuration¶
A VRF device is created with an associated route table. Network interfaces are then enslaved to a VRF device.
Configured routing table <id> is used by VRF <name>.
Note
A routing table ID can not be modified once it is assigned. It can only be changed by deleting and re-adding the VRF instance.
By default the scope of the port bindings for unbound sockets is limited to the default VRF. That is, it will not be matched by packets arriving on interfaces enslaved to a VRF and processes may bind to the same port if they bind to a VRF.
TCP & UDP services running in the default VRF context (ie., not bound to any VRF device) can work across all VRF domains by enabling this option.
Interfaces¶
When VRFs are used it is not only mandatory to create a VRF but also the VRF itself needs to be assigned to an interface.
Routing¶
Note
VyOS 1.4 (sagitta) introduced dynamic routing support for VRFs.
Currently dynamic routing is supported for the following protocols:
The CLI configuration is same as mentioned in above articles. The only difference is, that each routing protocol used, must be prefixed with the vrf name <name> command.
Example¶
The following commands would be required to set options for a given dynamic routing protocol inside a given vrf:
Operation¶
It is not sufficient to only configure a VRF but VRFs must be maintained, too. For VRF maintenance the following operational commands are in place.
Lists VRFs that have been created
vyos@vyos:~$ show vrf
VRF name state mac address flags interfaces
-------- ----- ----------- ----- ----------
blue up 00:53:12:d8:74:24 noarp,master,up,lower_up dum200,eth0.302
red up 00:53:de:02:df:aa noarp,master,up,lower_up dum100,eth0.300,bond0.100,peth0
Note
Command should probably be extended to list also the real interfaces assigned to this one VRF to get a better overview.
vyos@vyos:~$ show vrf name blue
VRF name state mac address flags interfaces
-------- ----- ----------- ----- ----------
blue up 00:53:12:d8:74:24 noarp,master,up,lower_up dum200,eth0.302
Display IPv4 routing table for VRF identified by <name>.
vyos@vyos:~$ show ip route vrf blue
Codes: K - kernel route, C - connected, S - static, R - RIP,
O - OSPF, I - IS-IS, B - BGP, E - EIGRP, N - NHRP,
T - Table, v - VNC, V - VNC-Direct, A - Babel, D - SHARP,
F - PBR, f - OpenFabric,
> - selected route, * - FIB route, q - queued route, r - rejected route
VRF blue:
K 0.0.0.0/0 [255/8192] unreachable (ICMP unreachable), 00:00:50
S>* 172.16.0.0/16 [1/0] via 192.0.2.1, dum1, 00:00:02
C>* 192.0.2.0/24 is directly connected, dum1, 00:00:06
Display IPv6 routing table for VRF identified by <name>.
vyos@vyos:~$ show ipv6 route vrf red
Codes: K - kernel route, C - connected, S - static, R - RIPng,
O - OSPFv3, I - IS-IS, B - BGP, N - NHRP, T - Table,
v - VNC, V - VNC-Direct, A - Babel, D - SHARP, F - PBR,
f - OpenFabric,
> - selected route, * - FIB route, q - queued route, r - rejected route
VRF red:
K ::/0 [255/8192] unreachable (ICMP unreachable), 00:43:20
C>* 2001:db8::/64 is directly connected, dum1, 00:02:19
C>* fe80::/64 is directly connected, dum1, 00:43:19
K>* ff00::/8 [0/256] is directly connected, dum1, 00:43:19
The ping command is used to test whether a network host is reachable or not.
Ping uses ICMP protocol’s mandatory ECHO_REQUEST datagram to elicit an ICMP ECHO_RESPONSE from a host or gateway. ECHO_REQUEST datagrams (pings) will have an IP and ICMP header, followed by “struct timeval” and an arbitrary number of pad bytes used to fill out the packet.
When doing fault isolation with ping, you should first run it on the local host, to verify that the local network interface is up and running. Then, continue with hosts and gateways further down the road towards your destination. Round-trip time and packet loss statistics are computed.
Duplicate packets are not included in the packet loss calculation, although the round-trip time of these packets is used in calculating the minimum/ average/maximum round-trip time numbers.
Ping command can be interrupted at any given time using <Ctrl>+c- A brief statistic is shown afterwards.
vyos@vyos:~$ ping 192.0.2.1 vrf red
PING 192.0.2.1 (192.0.2.1) 56(84) bytes of data.
64 bytes from 192.0.2.1: icmp_seq=1 ttl=64 time=0.070 ms
64 bytes from 192.0.2.1: icmp_seq=2 ttl=64 time=0.078 ms
^C
--- 192.0.2.1 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 4ms
rtt min/avg/max/mdev = 0.070/0.074/0.078/0.004 ms
Example¶
VRF route leaking¶
The following example topology was build using EVE-NG.
VRF route leaking
- PC1 is in the
defaultVRF and acting as e.g. a “fileserver” - PC2 is in VRF
bluewhich is the development department - PC3 and PC4 are connected to a bridge device on router
R1which is in VRFred. Say this is the HR department. - R1 is managed through an out-of-band network that resides in VRF
mgmt
Configuration¶
set interfaces bridge br10 address '10.30.0.254/24' set interfaces bridge br10 member interface eth3 set interfaces bridge br10 member interface eth4 set interfaces bridge br10 vrf 'red' set interfaces ethernet eth0 address 'dhcp' set interfaces ethernet eth0 vrf 'mgmt' set interfaces ethernet eth1 address '10.0.0.254/24' set interfaces ethernet eth2 address '10.20.0.254/24' set interfaces ethernet eth2 vrf 'blue' set protocols static route 10.20.0.0/24 interface eth2 vrf 'blue' set protocols static route 10.30.0.0/24 interface br10 vrf 'red' set service ssh disable-host-validation set service ssh vrf 'mgmt' set system name-servers-dhcp 'eth0' set vrf name blue protocols static route 10.0.0.0/24 interface eth1 vrf 'default' set vrf name blue table '3000' set vrf name mgmt table '1000' set vrf name red protocols static route 10.0.0.0/24 interface eth1 vrf 'default' set vrf name red table '2000'
Operation¶
After committing the configuration we can verify all leaked routes are installed, and try to ICMP ping PC1 from PC3.
PCS> ping 10.0.0.1 84 bytes from 10.0.0.1 icmp_seq=1 ttl=63 time=1.943 ms 84 bytes from 10.0.0.1 icmp_seq=2 ttl=63 time=1.618 ms 84 bytes from 10.0.0.1 icmp_seq=3 ttl=63 time=1.745 msVPCS> show ip NAME : VPCS[1] IP/MASK : 10.30.0.1/24 GATEWAY : 10.30.0.254 DNS : MAC : 00:50:79:66:68:0f
VRF default routing table¶
vyos@R1:~$ show ip route Codes: K - kernel route, C - connected, S - static, R - RIP, O - OSPF, I - IS-IS, B - BGP, E - EIGRP, N - NHRP, T - Table, v - VNC, V - VNC-Direct, A - Babel, D - SHARP, F - PBR, f - OpenFabric, > - selected route, * - FIB route, q - queued, r - rejected, b - backup C>* 10.0.0.0/24 is directly connected, eth1, 00:07:44 S>* 10.20.0.0/24 [1/0] is directly connected, eth2 (vrf blue), weight 1, 00:07:38 S>* 10.30.0.0/24 [1/0] is directly connected, br10 (vrf red), weight 1, 00:07:38
VRF red routing table¶
vyos@R1:~$ show ip route vrf red Codes: K - kernel route, C - connected, S - static, R - RIP, O - OSPF, I - IS-IS, B - BGP, E - EIGRP, N - NHRP, T - Table, v - VNC, V - VNC-Direct, A - Babel, D - SHARP, F - PBR, f - OpenFabric, > - selected route, * - FIB route, q - queued, r - rejected, b - backup VRF red: K>* 0.0.0.0/0 [255/8192] unreachable (ICMP unreachable), 00:07:57 S>* 10.0.0.0/24 [1/0] is directly connected, eth1 (vrf default), weight 1, 00:07:40 C>* 10.30.0.0/24 is directly connected, br10, 00:07:54
VRF blue routing table¶
vyos@R1:~$ show ip route vrf blue Codes: K - kernel route, C - connected, S - static, R - RIP, O - OSPF, I - IS-IS, B - BGP, E - EIGRP, N - NHRP, T - Table, v - VNC, V - VNC-Direct, A - Babel, D - SHARP, F - PBR, f - OpenFabric, > - selected route, * - FIB route, q - queued, r - rejected, b - backup VRF blue: K>* 0.0.0.0/0 [255/8192] unreachable (ICMP unreachable), 00:08:00 S>* 10.0.0.0/24 [1/0] is directly connected, eth1 (vrf default), weight 1, 00:07:44 C>* 10.20.0.0/24 is directly connected, eth2, 00:07:53