Documentation/networking/vrf.rst

   1 .. SPDX-License-Identifier: GPL-2.0
   2
   3 ====================================
   4 Virtual Routing and Forwarding (VRF)
   5 ====================================
   6
   7 The VRF Device
   8 ==============
   9
  10 The VRF device combined with ip rules provides the ability to create virtual
  11 routing and forwarding domains (aka VRFs, VRF-lite to be specific) in the
  12 Linux network stack. One use case is the multi-tenancy problem where each
  13 tenant has their own unique routing tables and in the very least need
  14 different default gateways.
  15
  16 Processes can be "VRF aware" by binding a socket to the VRF device. Packets
  17 through the socket then use the routing table associated with the VRF
  18 device. An important feature of the VRF device implementation is that it
  19 impacts only Layer 3 and above so L2 tools (e.g., LLDP) are not affected
  20 (ie., they do not need to be run in each VRF). The design also allows
  21 the use of higher priority ip rules (Policy Based Routing, PBR) to take
  22 precedence over the VRF device rules directing specific traffic as desired.
  23
  24 In addition, VRF devices allow VRFs to be nested within namespaces. For
  25 example network namespaces provide separation of network interfaces at the
  26 device layer, VLANs on the interfaces within a namespace provide L2 separation
  27 and then VRF devices provide L3 separation.
  28
  29 Design
  30 ------
  31 A VRF device is created with an associated route table. Network interfaces
  32 are then enslaved to a VRF device::
  33
  34          +-----------------------------+
  35          |           vrf-blue          |  ===> route table 10
  36          +-----------------------------+
  37             |        |            |
  38          +------+ +------+     +-------------+
  39          | eth1 | | eth2 | ... |    bond1    |
  40          +------+ +------+     +-------------+
  41                                   |       |
  42                               +------+ +------+
  43                               | eth8 | | eth9 |
  44                               +------+ +------+
  45
  46 Packets received on an enslaved device and are switched to the VRF device
  47 in the IPv4 and IPv6 processing stacks giving the impression that packets
  48 flow through the VRF device. Similarly on egress routing rules are used to
  49 send packets to the VRF device driver before getting sent out the actual
  50 interface. This allows tcpdump on a VRF device to capture all packets into
  51 and out of the VRF as a whole\ [1]_. Similarly, netfilter\ [2]_ and tc rules
  52 can be applied using the VRF device to specify rules that apply to the VRF
  53 domain as a whole.
  54
  55 .. [1] Packets in the forwarded state do not flow through the device, so those
  56        packets are not seen by tcpdump. Will revisit this limitation in a
  57        future release.
  58
  59 .. [2] Iptables on ingress supports PREROUTING with skb->dev set to the real
  60        ingress device and both INPUT and PREROUTING rules with skb->dev set to
  61        the VRF device. For egress POSTROUTING and OUTPUT rules can be written
  62        using either the VRF device or real egress device.
  63
  64 Setup
  65 -----
  66 1. VRF device is created with an association to a FIB table.
  67    e.g,::
  68
  69         ip link add vrf-blue type vrf table 10
  70         ip link set dev vrf-blue up
  71
  72 2. An l3mdev FIB rule directs lookups to the table associated with the device.
  73    A single l3mdev rule is sufficient for all VRFs. The VRF device adds the
  74    l3mdev rule for IPv4 and IPv6 when the first device is created with a
  75    default preference of 1000. Users may delete the rule if desired and add
  76    with a different priority or install per-VRF rules.
  77
  78    Prior to the v4.8 kernel iif and oif rules are needed for each VRF device::
  79
  80        ip ru add oif vrf-blue table 10
  81        ip ru add iif vrf-blue table 10
  82
  83 3. Set the default route for the table (and hence default route for the VRF)::
  84
  85        ip route add table 10 unreachable default metric 4278198272
  86
  87    This high metric value ensures that the default unreachable route can
  88    be overridden by a routing protocol suite.  FRRouting interprets
  89    kernel metrics as a combined admin distance (upper byte) and priority
  90    (lower 3 bytes).  Thus the above metric translates to [255/8192].
  91
  92 4. Enslave L3 interfaces to a VRF device::
  93
  94        ip link set dev eth1 master vrf-blue
  95
  96    Local and connected routes for enslaved devices are automatically moved to
  97    the table associated with VRF device. Any additional routes depending on
  98    the enslaved device are dropped and will need to be reinserted to the VRF
  99    FIB table following the enslavement.
 100
 101    The IPv6 sysctl option keep_addr_on_down can be enabled to keep IPv6 global
 102    addresses as VRF enslavement changes::
 103
 104        sysctl -w net.ipv6.conf.all.keep_addr_on_down=1
 105
 106 5. Additional VRF routes are added to associated table::
 107
 108        ip route add table 10 ...
 109
 110
 111 Applications
 112 ------------
 113 Applications that are to work within a VRF need to bind their socket to the
 114 VRF device::
 115
 116     setsockopt(sd, SOL_SOCKET, SO_BINDTODEVICE, dev, strlen(dev)+1);
 117
 118 or to specify the output device using cmsg and IP_PKTINFO.
 119
 120 By default the scope of the port bindings for unbound sockets is
 121 limited to the default VRF. That is, it will not be matched by packets
 122 arriving on interfaces enslaved to an l3mdev and processes may bind to
 123 the same port if they bind to an l3mdev.
 124
 125 TCP & UDP services running in the default VRF context (ie., not bound
 126 to any VRF device) can work across all VRF domains by enabling the
 127 tcp_l3mdev_accept and udp_l3mdev_accept sysctl options::
 128
 129     sysctl -w net.ipv4.tcp_l3mdev_accept=1
 130     sysctl -w net.ipv4.udp_l3mdev_accept=1
 131
 132 These options are disabled by default so that a socket in a VRF is only
 133 selected for packets in that VRF. There is a similar option for RAW
 134 sockets, which is enabled by default for reasons of backwards compatibility.
 135 This is so as to specify the output device with cmsg and IP_PKTINFO, but
 136 using a socket not bound to the corresponding VRF. This allows e.g. older ping
 137 implementations to be run with specifying the device but without executing it
 138 in the VRF. This option can be disabled so that packets received in a VRF
 139 context are only handled by a raw socket bound to the VRF, and packets in the
 140 default VRF are only handled by a socket not bound to any VRF::
 141
 142     sysctl -w net.ipv4.raw_l3mdev_accept=0
 143
 144 netfilter rules on the VRF device can be used to limit access to services
 145 running in the default VRF context as well.
 146
 147 --------------------------------------------------------------------------------
 148
 149 Using iproute2 for VRFs
 150 =======================
 151 iproute2 supports the vrf keyword as of v4.7. For backwards compatibility this
 152 section lists both commands where appropriate -- with the vrf keyword and the
 153 older form without it.
 154
 155 1. Create a VRF
 156
 157    To instantiate a VRF device and associate it with a table::
 158
 159        $ ip link add dev NAME type vrf table ID
 160
 161    As of v4.8 the kernel supports the l3mdev FIB rule where a single rule
 162    covers all VRFs. The l3mdev rule is created for IPv4 and IPv6 on first
 163    device create.
 164
 165 2. List VRFs
 166
 167    To list VRFs that have been created::
 168
 169        $ ip [-d] link show type vrf
 170          NOTE: The -d option is needed to show the table id
 171
 172    For example::
 173
 174        $ ip -d link show type vrf
 175        11: mgmt: <NOARP,MASTER,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DEFAULT group default qlen 1000
 176            link/ether 72:b3:ba:91:e2:24 brd ff:ff:ff:ff:ff:ff promiscuity 0
 177            vrf table 1 addrgenmode eui64
 178        12: red: <NOARP,MASTER,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DEFAULT group default qlen 1000
 179            link/ether b6:6f:6e:f6:da:73 brd ff:ff:ff:ff:ff:ff promiscuity 0
 180            vrf table 10 addrgenmode eui64
 181        13: blue: <NOARP,MASTER,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DEFAULT group default qlen 1000
 182            link/ether 36:62:e8:7d:bb:8c brd ff:ff:ff:ff:ff:ff promiscuity 0
 183            vrf table 66 addrgenmode eui64
 184        14: green: <NOARP,MASTER,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DEFAULT group default qlen 1000
 185            link/ether e6:28:b8:63:70:bb brd ff:ff:ff:ff:ff:ff promiscuity 0
 186            vrf table 81 addrgenmode eui64
 187
 188
 189    Or in brief output::
 190
 191        $ ip -br link show type vrf
 192        mgmt         UP             72:b3:ba:91:e2:24 <NOARP,MASTER,UP,LOWER_UP>
 193        red          UP             b6:6f:6e:f6:da:73 <NOARP,MASTER,UP,LOWER_UP>
 194        blue         UP             36:62:e8:7d:bb:8c <NOARP,MASTER,UP,LOWER_UP>
 195        green        UP             e6:28:b8:63:70:bb <NOARP,MASTER,UP,LOWER_UP>
 196
 197
 198 3. Assign a Network Interface to a VRF
 199
 200    Network interfaces are assigned to a VRF by enslaving the netdevice to a
 201    VRF device::
 202
 203        $ ip link set dev NAME master NAME
 204
 205    On enslavement connected and local routes are automatically moved to the
 206    table associated with the VRF device.
 207
 208    For example::
 209
 210        $ ip link set dev eth0 master mgmt
 211
 212
 213 4. Show Devices Assigned to a VRF
 214
 215    To show devices that have been assigned to a specific VRF add the master
 216    option to the ip command::
 217
 218        $ ip link show vrf NAME
 219        $ ip link show master NAME
 220
 221    For example::
 222
 223        $ ip link show vrf red
 224        3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master red state UP mode DEFAULT group default qlen 1000
 225            link/ether 02:00:00:00:02:02 brd ff:ff:ff:ff:ff:ff
 226        4: eth2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master red state UP mode DEFAULT group default qlen 1000
 227            link/ether 02:00:00:00:02:03 brd ff:ff:ff:ff:ff:ff
 228        7: eth5: <BROADCAST,MULTICAST> mtu 1500 qdisc noop master red state DOWN mode DEFAULT group default qlen 1000
 229            link/ether 02:00:00:00:02:06 brd ff:ff:ff:ff:ff:ff
 230
 231
 232    Or using the brief output::
 233
 234        $ ip -br link show vrf red
 235        eth1             UP             02:00:00:00:02:02 <BROADCAST,MULTICAST,UP,LOWER_UP>
 236        eth2             UP             02:00:00:00:02:03 <BROADCAST,MULTICAST,UP,LOWER_UP>
 237        eth5             DOWN           02:00:00:00:02:06 <BROADCAST,MULTICAST>
 238
 239
 240 5. Show Neighbor Entries for a VRF
 241
 242    To list neighbor entries associated with devices enslaved to a VRF device
 243    add the master option to the ip command::
 244
 245        $ ip [-6] neigh show vrf NAME
 246        $ ip [-6] neigh show master NAME
 247
 248    For example::
 249
 250        $  ip neigh show vrf red
 251        10.2.1.254 dev eth1 lladdr a6:d9:c7:4f:06:23 REACHABLE
 252        10.2.2.254 dev eth2 lladdr 5e:54:01:6a:ee:80 REACHABLE
 253
 254        $ ip -6 neigh show vrf red
 255        2002:1::64 dev eth1 lladdr a6:d9:c7:4f:06:23 REACHABLE
 256
 257
 258 6. Show Addresses for a VRF
 259
 260    To show addresses for interfaces associated with a VRF add the master
 261    option to the ip command::
 262
 263        $ ip addr show vrf NAME
 264        $ ip addr show master NAME
 265
 266    For example::
 267
 268         $ ip addr show vrf red
 269         3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master red state UP group default qlen 1000
 270             link/ether 02:00:00:00:02:02 brd ff:ff:ff:ff:ff:ff
 271             inet 10.2.1.2/24 brd 10.2.1.255 scope global eth1
 272                valid_lft forever preferred_lft forever
 273             inet6 2002:1::2/120 scope global
 274                valid_lft forever preferred_lft forever
 275             inet6 fe80::ff:fe00:202/64 scope link
 276                valid_lft forever preferred_lft forever
 277         4: eth2: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master red state UP group default qlen 1000
 278             link/ether 02:00:00:00:02:03 brd ff:ff:ff:ff:ff:ff
 279             inet 10.2.2.2/24 brd 10.2.2.255 scope global eth2
 280                valid_lft forever preferred_lft forever
 281             inet6 2002:2::2/120 scope global
 282                valid_lft forever preferred_lft forever
 283             inet6 fe80::ff:fe00:203/64 scope link
 284                valid_lft forever preferred_lft forever
 285         7: eth5: <BROADCAST,MULTICAST> mtu 1500 qdisc noop master red state DOWN group default qlen 1000
 286             link/ether 02:00:00:00:02:06 brd ff:ff:ff:ff:ff:ff
 287
 288    Or in brief format::
 289
 290         $ ip -br addr show vrf red
 291         eth1             UP             10.2.1.2/24 2002:1::2/120 fe80::ff:fe00:202/64
 292         eth2             UP             10.2.2.2/24 2002:2::2/120 fe80::ff:fe00:203/64
 293         eth5             DOWN
 294
 295
 296 7. Show Routes for a VRF
 297
 298    To show routes for a VRF use the ip command to display the table associated
 299    with the VRF device::
 300
 301        $ ip [-6] route show vrf NAME
 302        $ ip [-6] route show table ID
 303
 304    For example::
 305
 306         $ ip route show vrf red
 307         unreachable default  metric 4278198272
 308         broadcast 10.2.1.0 dev eth1  proto kernel  scope link  src 10.2.1.2
 309         10.2.1.0/24 dev eth1  proto kernel  scope link  src 10.2.1.2
 310         local 10.2.1.2 dev eth1  proto kernel  scope host  src 10.2.1.2
 311         broadcast 10.2.1.255 dev eth1  proto kernel  scope link  src 10.2.1.2
 312         broadcast 10.2.2.0 dev eth2  proto kernel  scope link  src 10.2.2.2
 313         10.2.2.0/24 dev eth2  proto kernel  scope link  src 10.2.2.2
 314         local 10.2.2.2 dev eth2  proto kernel  scope host  src 10.2.2.2
 315         broadcast 10.2.2.255 dev eth2  proto kernel  scope link  src 10.2.2.2
 316
 317         $ ip -6 route show vrf red
 318         local 2002:1:: dev lo  proto none  metric 0  pref medium
 319         local 2002:1::2 dev lo  proto none  metric 0  pref medium
 320         2002:1::/120 dev eth1  proto kernel  metric 256  pref medium
 321         local 2002:2:: dev lo  proto none  metric 0  pref medium
 322         local 2002:2::2 dev lo  proto none  metric 0  pref medium
 323         2002:2::/120 dev eth2  proto kernel  metric 256  pref medium
 324         local fe80:: dev lo  proto none  metric 0  pref medium
 325         local fe80:: dev lo  proto none  metric 0  pref medium
 326         local fe80::ff:fe00:202 dev lo  proto none  metric 0  pref medium
 327         local fe80::ff:fe00:203 dev lo  proto none  metric 0  pref medium
 328         fe80::/64 dev eth1  proto kernel  metric 256  pref medium
 329         fe80::/64 dev eth2  proto kernel  metric 256  pref medium
 330         ff00::/8 dev red  metric 256  pref medium
 331         ff00::/8 dev eth1  metric 256  pref medium
 332         ff00::/8 dev eth2  metric 256  pref medium
 333         unreachable default dev lo  metric 4278198272  error -101 pref medium
 334
 335 8. Route Lookup for a VRF
 336
 337    A test route lookup can be done for a VRF::
 338
 339        $ ip [-6] route get vrf NAME ADDRESS
 340        $ ip [-6] route get oif NAME ADDRESS
 341
 342    For example::
 343
 344         $ ip route get 10.2.1.40 vrf red
 345         10.2.1.40 dev eth1  table red  src 10.2.1.2
 346             cache
 347
 348         $ ip -6 route get 2002:1::32 vrf red
 349         2002:1::32 from :: dev eth1  table red  proto kernel  src 2002:1::2  metric 256  pref medium
 350
 351
 352 9. Removing Network Interface from a VRF
 353
 354    Network interfaces are removed from a VRF by breaking the enslavement to
 355    the VRF device::
 356
 357        $ ip link set dev NAME nomaster
 358
 359    Connected routes are moved back to the default table and local entries are
 360    moved to the local table.
 361
 362    For example::
 363
 364     $ ip link set dev eth0 nomaster
 365
 366 --------------------------------------------------------------------------------
 367
 368 Commands used in this example::
 369
 370      cat >> /etc/iproute2/rt_tables.d/vrf.conf <<EOF
 371      1  mgmt
 372      10 red
 373      66 blue
 374      81 green
 375      EOF
 376
 377      function vrf_create
 378      {
 379          VRF=$1
 380          TBID=$2
 381
 382          # create VRF device
 383          ip link add ${VRF} type vrf table ${TBID}
 384
 385          if [ "${VRF}" != "mgmt" ]; then
 386              ip route add table ${TBID} unreachable default metric 4278198272
 387          fi
 388          ip link set dev ${VRF} up
 389      }
 390
 391      vrf_create mgmt 1
 392      ip link set dev eth0 master mgmt
 393
 394      vrf_create red 10
 395      ip link set dev eth1 master red
 396      ip link set dev eth2 master red
 397      ip link set dev eth5 master red
 398
 399      vrf_create blue 66
 400      ip link set dev eth3 master blue
 401
 402      vrf_create green 81
 403      ip link set dev eth4 master green
 404
 405
 406      Interface addresses from /etc/network/interfaces:
 407      auto eth0
 408      iface eth0 inet static
 409            address 10.0.0.2
 410            netmask 255.255.255.0
 411            gateway 10.0.0.254
 412
 413      iface eth0 inet6 static
 414            address 2000:1::2
 415            netmask 120
 416
 417      auto eth1
 418      iface eth1 inet static
 419            address 10.2.1.2
 420            netmask 255.255.255.0
 421
 422      iface eth1 inet6 static
 423            address 2002:1::2
 424            netmask 120
 425
 426      auto eth2
 427      iface eth2 inet static
 428            address 10.2.2.2
 429            netmask 255.255.255.0
 430
 431      iface eth2 inet6 static
 432            address 2002:2::2
 433            netmask 120
 434
 435      auto eth3
 436      iface eth3 inet static
 437            address 10.2.3.2
 438            netmask 255.255.255.0
 439
 440      iface eth3 inet6 static
 441            address 2002:3::2
 442            netmask 120
 443
 444      auto eth4
 445      iface eth4 inet static
 446            address 10.2.4.2
 447            netmask 255.255.255.0
 448
 449      iface eth4 inet6 static
 450            address 2002:4::2
 451            netmask 120