Example: VPLS Configuration (BGP Signaling)
In Figure 1, a simple VPLS topology is enabled between routers PE1 and PE2. CE routers CE1 and CE2 use Ethernet-based interfaces to connect VLAN 600 to their local PE router. The PE routers PE1 and PE2 are connected to one another by LSPs enabled across a service provider backbone running MPLS, BGP, RSVP, and OSPF.
In a VPLS routing instance named green, PE1 has a local interface fe-0/1/0 and a virtual port
of vt-0/3/0.32770 (the virtual port is created dynamically
on the Tunnel Services PIC when VPLS is configured). PE2 has a local
interface fe-0/1/0 and a virtual port of vt-0/3/0.32771 in the same green instance. As a result, routers CE1
and CE2 send Ethernet traffic to one another as if they were physically
connected to each other on a LAN.
On Router CE1, the only item you need to configure is the Fast Ethernet interface that connects to PE1. Be sure to write down the VLAN identifier and IP address, so you can match them later on CE2.
Router CE1
[edit]
interfaces {
fe-0/1/0 {
vlan-tagging; # Configure VLAN tagging for VLAN VPLS or extended VLAN VPLS.
unit 0 {
vlan-id 600; # The Ethernet interface on CE2 must use the same VLAN ID.
family inet {
address 10.11.3.1/24; # The interface on CE2 must use the same prefix.
}
}
}
}
If Router PE1 is an MX Series device, you need to configure a tunnel service interface.
To create tunnel interfaces on an MX Series
router, include the tunnel-services statement at the [edit chassis fpc slot-number pic number] hierarchy level. To configure the bandwidth for a tunnel
interface, include the bandwidth statement at the [edit chassis fpc slot-number pic number tunnel services] hierarchy level.
The following example shows a tunnel interface with 1 Gbps of bandwidth configured on PFE 3 of the DPC installed in slot 0 of an MX Series router:
[edit chassis]
fpc 0 {
pic 3 {
tunnel services {
bandwidth 1g;
}
}
}
On Router PE1, prepare the router for VPLS by configuring BGP,
MPLS, OSPF, and RSVP. (These protocols are the basis for most Layer 2
VPN-related applications, including VPLS.) Include the signaling statement at the [edit protocols bgp group group-name family l2vpn] hierarchy level, because VPLS uses the same
infrastructure for internal BGP as Layer 2 VPNs.
In Junos OS Release 7.3 and later, the signaling statement replaces the unicast statement
at the [edit protocols bgp group group-name family l2vpn] hierarchy level. You must use the signaling statement if you wish to configure VPLS domains and Layer 2
VPNs simultaneously.
Next, configure VLAN tagging on the Fast Ethernet interface connected to Router CE1. Include VLAN VPLS encapsulation at both the physical and logical interface levels. Be sure to use the same VLAN ID for all Ethernet interfaces that are part of a single VPLS instance. Finally, add the Fast Ethernet interface into a VPLS routing instance and specify the site range, site ID number, and site name.
Router PE1
[edit]
interfaces {
fe-0/1/0 {
vlan-tagging;# Configure VLAN tagging for VLAN VPLS or extended VLAN VPLS.
encapsulation vlan-vpls; # Configure VPLS encapsulation on both the
unit 0 { # physical interface and the logical interface.
encapsulation vlan-vpls;
vlan-id 600;# The VLAN ID is the same one used by the CE routers.
}
}
so-1/1/0 {
unit 0 {
family inet {
address 10.11.1.5/30;
}
family mpls;
}
}
lo0 {
unit 0 {
family inet {
address 10.245.14.218/32;
}
}
}
}
routing-options {
autonomous-system 69;
forwarding-table {
export exp-to-fwd; # Apply a policy that selects an LSP for the VPLS instance.
}
}
protocols {
rsvp {
interface all {
aggregate;
}
}
mpls {
label-switched-path pe1-to-pe2 { # Configure an LSP to reach other VPLS PEs.
to 10.245.14.219;
}
interface all;
}
bgp {
group vpls-pe {
type internal;
local-address 10.245.14.218;
family l2vpn { # VPLS uses the same infrastructure as Layer 2 VPNs
signaling; # for internal BGP.
}
neighbor 10.245.14.217;
neighbor 10.245.14.219;
}
}
ospf {
traffic-engineering;
area 0.0.0.0 {
interface so-1/1/0.0 {
metric 11;
}
interface lo0.0 {
passive;
}
}
}
}
policy-options {
policy-statement exp-to-fwd {
term a {
from community grn-com; # Matches the community in the VPLS instance.
then {
install-nexthop lsp pe1-to-pe2; # If there are multiple LSPs that exist
accept; # between VPLS PE routers, this statement sends VPLS traffic
} # over a specific LSP.
}
}
community grn-com members target:11111:1; # Adds the instance to a BGP
} # community.
routing-instances {
green {
instance-type vpls; # Configure a VPLS routing instance.
interface fe-0/1/0.0;
route-distinguisher 10.245.14.218:1;
vrf-target target:11111:1; # This value is important to the BGP community.
protocols {
vpls { # Configure a VPLS site range, site name, and site identifier.
site-range 10;
site greenPE1 {
site-identifier 1;
}
}
}
}
}
On Router P0, configure BGP, MPLS, OSPF, and RSVP to interconnect PE1 and PE2.
Router P0
[edit]
interfaces {
so-0/0/0 {
unit 0 {
family inet {
address 10.11.2.6/30;
}
family mpls;
}
}
so-1/1/0 {
unit 0 {
family inet {
address 10.11.1.6/30;
}
family mpls;
}
}
lo0 {
unit 0 {
family inet {
address 10.245.14.217/32;
}
}
}
}
routing-options {
autonomous-system 69;
}
protocols {
rsvp {
interface all {
aggregate;
}
}
mpls {
interface all;
}
bgp {
group vpls-pe {
type internal;
local-address 10.245.14.217;
family l2vpn { # VPLS uses the same infrastructure as Layer 2 VPNs
signaling; # for internal BGP.
}
neighbor 10.245.14.218;
neighbor 10.245.14.219;
}
}
ospf {
traffic-engineering;
area 0.0.0.0 {
interface so-1/1/0.0 {
metric 11;
}
interface so-0/0/0.0 {
metric 15;
}
interface lo0.0 {
passive;
}
}
}
}
If Router PE2 is an MX Series device, you need to configure a tunnel service interfaces.
To create tunnel interfaces on an MX Series
router, include the tunnel-services statement at the [edit chassis fpc slot-number pic number] hierarchy level. To configure the bandwidth for a tunnel
interface, include the bandwidth statement at the [edit chassis fpc slot-number pic number] hierarchy level.
The following example shows a tunnel interface with 1 Gbps of bandwidth configured on PFE 3 of the DPC installed in slot 0 of an MX Series router:
[edit chassis]
fpc 0 {
pic 3 {
tunnel services {
bandwidth 1g;
}
}
}
On Router PE2, configure BGP, MPLS, OSPF, and RSVP to complement the configuration on PE1. Next, configure VLAN tagging on the Fast Ethernet interface connected to Router CE2. Include VLAN VPLS encapsulation at both the physical and logical interface levels. Be sure to use the same VLAN ID for all Ethernet interfaces that are part of a single VPLS instance. Finally, add the Fast Ethernet interface into a VPLS routing instance and specify the site range, site ID number, and site name.
Router PE2
[edit]
interfaces {
fe-0/1/0 {
vlan-tagging; # Configure VLAN tagging for VLAN VPLS or extended VLAN VPLS.
encapsulation vlan-vpls; # Configure VPLS encapsulation on both the
unit 0 { # physical interface and logical interface.
encapsulation vlan-vpls;
vlan-id 600;# The VLAN ID is the same one used by the CE routers.
}
}
so-0/0/0 {
unit 0 {
family inet {
address 10.11.2.5/30;
}
family mpls;
}
}
lo0 {
unit 0 {
family inet {
address 10.245.14.219/32;
}
}
}
}
routing-options {
autonomous-system 69;
forwarding-table {
export exp-to-fwd; # Apply a policy that selects an LSP for the VPLS instance.
}
}
protocols {
rsvp {
interface all {
aggregate;
}
}
mpls {
label-switched-path pe2-to-pe1 { # Configure an LSP to other VPLS PE routers.
to 10.245.14.218;
}
interface all;
}
bgp {
group vpls-pe {
type internal;
local-address 10.245.14.219;
family l2vpn { # VPLS uses the same infrastructure as Layer 2 VPNs
signaling; # for internal BGP.
}
neighbor 10.245.14.217;
neighbor 10.245.14.218;
}
}
ospf {
traffic-engineering;
area 0.0.0.0 {
interface so-0/0/0.0 {
metric 15;
}
interface lo0.0 {
passive;
}
}
}
}
policy-options {
policy-statement exp-to-fwd {
term a {
from community grn-com; # Matches the community with the VPLS instance.
then {
install-nexthop lsp pe2-to-pe1; # If there are multiple LSPs that exist
accept; # between VPLS PE routers, this statement sends VPLS traffic
} # over a specific LSP.
}
}
community grn-com members target:11111:1; # This adds the instance into a BGP community.
}
routing-instances {
green {
instance-type vpls; # Configure a VPLS routing instance.
interface fe-0/1/0.0;
route-distinguisher 10.245.14.219:1;
vrf-target target:11111:1; # This value is important for the BGP community.
protocols {
vpls { # Configure a VPLS site range, site name, and site identifier.
site-range 10;
site greenPE2 {
site-identifier 2;
}
}
}
}
}
On Router CE2, complete your VPLS network by configuring the Fast Ethernet interface that connects to PE2. Use the same VLAN identifier and IP address prefix used on Router CE1.
Router CE2
[edit]
interfaces {
fe-0/1/0 {
vlan-tagging; # Configure VLAN tagging for VLAN VPLS or extended VLAN VPLS.
unit 0 {
vlan-id 600; # The Ethernet interface on CE1 must use the same VLAN ID.
family inet {
address 10.11.3.2/24; # The interface on CE1 must use the same prefix.
}
}
}
}
Verifying Your Work
To verify proper operation of VPLS, use the following commands:
clear vpls mac-address instance instance-nameshow interfaces terseshow route forwarding-table family mplsshow route forwarding-table family vpls (destination | extensive | matching | table)show route instance (detail)show system statistics vplsshow vpls connectionsshow vpls statistics
The following section shows the output of these commands on Router PE1 as a result of the configuration example:
user@PE1> show interfaces terse
Interface Admin Link Proto Local Remote
so-1/1/0 up up
so-1/1/0.0 up up inet 10.11.1.5/30
mpls
so-1/1/1 up up
so-1/1/2 up up
so-1/1/3 up up
fe-0/1/0 up up
fe-0/1/0.0 up up vpls # This is the local Fast Ethernet
# interface.
fe-0/1/1 up up
fe-0/1/2 up up
fe-0/1/3 up up
gr-0/3/0 up up
ip-0/3/0 up up
mt-0/3/0 up up
pd-0/3/0 up up
pe-0/3/0 up up
vt-0/3/0 up up
vt-0/3/0.32770 up up # This is the dynamically generated virtual port.
dsc up up
fxp0 up up
fxp0.0 up up inet 192.186.14.218/24
fxp1 up up
fxp1.0 up up tnp 4
gre up up
ipip up up
lo0 up up
lo0.0 up up inet 10.245.14.218 --> 0/0
127.0.0.1 --> 0/0
inet6 fe80::2a0:a5ff:fe28:13e0
feee::10:245:14:218
lsi up up
mtun up up
pimd up up
pime up up
tap up up
user@PE1> show system statistics vpls
vpls:
0 total packets received
0 with size smaller than minimum
0 with incorrect version number
0 packets for this host
0 packets with no logical interface
0 packets with no family
0 packets with no route table
0 packets with no auxiliary table
0 packets with no corefacing entry
0 packets with no CE-facing entry
6 mac route learning requests # This indicates that VPLS is working.
6 mac routes learnt
0 mac routes aged
0 mac routes moved
To display VPLS source and destination MAC address
accounting information, use the destination, extensive, matching, or table option with the show
route forwarding-table family vpls command. When you analyze
the display output, keep in mind the following:
VPLS MAC address accounting is handled on a per-MAC address basis for each VPLS instance. All information is retrieved from MAC address entries in the MAC address table. VPLS MAC address accounting is performed only on local CE routers.
The VPLS counters for source and destination MAC addresses increment continuously until the oldest MAC address entries are removed from the memory buffer, either when the entries time out or if the VPLS instance is restarted.
user@PE1> show route forwarding-table family vpls extensive Routing table: green.vpls [Index 2] VPLS: Destination: default Route type: dynamic Route reference: 0 Flags: sent to PFE Next-hop type: flood Index: 353 Reference: 1 Destination: default Route type: permanent Route reference: 0 Flags: none Next-hop type: discard Index: 298 Reference: 1 Destination: fe-0/1/0.0 Route type: dynamic Route reference: 0 Flags: sent to PFE Next-hop type: flood Index: 355 Reference: 1 Destination: bb:bb:bb:bb:bb:bb/48 # This MAC address belongs to remote CE2. Route type: dynamic Route reference: 0 Flags: sent to PFE, prefix load balance Next-hop type: indirect Index: 351 Reference: 4 Next-hop type: Push 800000, Push 100002(top) Next-hop interface: so-1/1/0.0 Destination: aa:aa:aa:aa:aa:aa/48 # This MAC address belongs to local CE1. Route type: dynamic Route reference: 0 Flags: sent to PFE, prefix load balance Next-hop type: unicast Index: 354 Reference: 2 Next-hop interface: fe-0/1/0.0 user@PE1> show route forwarding-table family vpls Routing table: green.vpls VPLS: Destination Type RtRef Next hop Type Index NhRef Netif default dynm 0 flood 353 1 default perm 0 dscd 298 1 fe-0/1/0.0 dynm 0 flood 355 1 bb:bb:bb:bb:bb:bb/48 # This MAC address belongs to remote CE2. dynm 0 indr 351 4 Push 800000, Push 100002(top) so-1/1/0.0 aa:aa:aa:aa:aa:aa/48 # This MAC address belongs to local CE1. dynm 0 ucst 354 2 fe-0/1/0.0 user@PE1> show route forwarding-table family mpls Routing table: mpls MPLS: Destination Type RtRef Next hop Type Index NhRef Netif default perm 0 dscd 19 1 0 user 0 recv 18 3 1 user 0 recv 18 3 2 user 0 recv 18 3 100000 user 0 10.11.1.6 swap 100001 so-1/1/0.0 800002 user 0 Pop vt-0/3/0.32770 vt-0/3/0.32770 (VPLS) user 0 indr 351 4 Push 800000, Push 100002(top) so-1/1/0.0 user@PE1> show route instance green detail green: Router ID: 0.0.0.0 Type: vpls State: Active Interfaces: fe-0/1/0.0 # This is the local Fast Ethernet interface. vt-0/3/0.32770 # This is the dynamically generated VPLS virtual port. Route-distinguisher: 10.245.14.218:1 Vrf-import: [ __vrf-import-green-internal__ ] Vrf-export: [ __vrf-export-green-internal__ ] Vrf-import-target: [ target:11111:1 ] Vrf-export-target: [ target:11111:1 ] Tables: green.l2vpn.0 : 2 routes (2 active, 0 holddown, 0 hidden) user@PE1> show vpls connections L2VPN Connections: Legend for connection status (St) OR -- out of range WE -- intf encaps != instance encaps EI -- encapsulation invalid Dn -- down EM -- encapsulation mismatch VC-Dn -- Virtual circuit down CM -- control-word mismatch -> -- only outbound conn is up CN -- circuit not present <- -- only inbound conn is up OL -- no outgoing label Up -- operational NC -- intf encaps not CCC/TCC XX -- unknown NP -- interface not present Legend for interface status Up -- operational Dn -- down Instance: green Local site: greenPE1 (1) connection-site Type St Time last up # Up trans 2 rmt Up Jan 24 06:26:49 2003 1 Local interface: vt-0/3/0.32770, Status: Up, Encapsulation: VPLS Remote PE: 10.245.14.219, Negotiated control-word: No Incoming label: 800002, Outgoing label: 800000 user@PE1> show system statistics vpls vpls: 0 total packets received 0 with size smaller than minimum 0 with incorrect version number 0 packets for this host 0 packets with no logical interface 0 packets with no family 0 packets with no route table 0 packets with no auxiliary table 0 packets with no corefacing entry 0 packets with no CE-facing entry 7 mac route learning requests 7 mac routes learnt 0 mac routes aged 0 mac routes moved user@PE1> show route instance green detail green: Router ID: 0.0.0.0 Type: vpls State: Active Interfaces: fe-0/1/0.0 vt-0/3/0.32770 Route-distinguisher: 10.245.14.218:1 Vrf-import: [ __vrf-import-green-internal__ ] Vrf-export: [ __vrf-export-green-internal__ ] Vrf-import-target: [ target:11111:1 ] Vrf-export-target: [ target:11111:1 ] Tables: green.l2vpn.0 : 2 routes (2 active, 0 holddown, 0 hidden) user@PE1> show vpls statistics Layer-2 VPN Statistics: Instance: green Local interface: fe-0/1/0.0, Index: 351 Remote provider edge router: 10.245.14.219 Multicast packets: 363 Multicast bytes : 30956 Flood packets : 0 Flood bytes : 0 Local interface: vt-0/3/0.32770, Index: 354 Remote provider edge router: 10.245.14.219 Multicast packets: 135 Multicast bytes : 12014 Flood packets : 135 Flood bytes : 12014
To clear all MAC address entries for a VPLS instance
from the VPLS table, issue the clear vpls mac-address instance instance-name command. Add the logical-system logical-system-name option to clear entries in a
VPLS instance within a logical system. Use the mac-address option to remove individual MAC addresses.