G-SID Compression-Based SRv6 Inter-AS Option B Configuration

Introduction

In cross-data-center or cross-carrier networks, each network is typically managed as an independent autonomous system (AS). Inter-AS services require not only the exchange of VPN routes between ASs, but also that service traffic is forwarded along the intended path. SRv6 carries forwarding instructions in IPv6 addresses, providing a unified transport for inter-AS L3VPN services and supporting control over the transport path.

This case presents an SRv6 inter-AS networking solution that combines Inter-AS Option B, G-SID, and SR-TE Policy. VPN routes are exchanged between ASs using Option B; the ingress PE selects the inter-AS path based on the SR-TE Policy associated with the service route; and G-SID compresses the SIDs in the path to reduce packet encapsulation overhead.

Technical Principles

Inter-AS Option B Solution

In the Inter-AS Option B solution, each AS maintains its own intra-AS underlay network independently, and ASBRs exchange VPN routes via MP-eBGP so that service sites in different ASs can communicate. Inter-AS VPN routes are passed through the ASBRs to the ingress PE, which performs SRv6 encapsulation on service packets based on the route information; the egress PE decapsulates the packets and forwards them to the destination VPN.

G-SID Encapsulation

In traditional SRv6, each SID occupies 128 bits. As the number of segments in a path grows, the SID list increases packet header overhead.

G-SID uses the address portion shared by multiple nodes as the Common Block, and uses shorter SIDs to represent individual nodes or forwarding behaviors. Four G-SIDs (4 x 32 bits) can be encoded in a single IPv6 address (128 bits). When a packet passes through a node, the node processes the current SID and updates the packet’s IPv6 destination address so that the packet continues to be forwarded along the specified path.

In this case, G-SID compressed encoding is used in the Segment List of the SR-TE Policy to reduce the encapsulation overhead of the inter-AS explicit path.

Typical Configuration Example

Network Topology

IP Address Plan

Link	Local End	IP	Remote End	IP
CE1-PE1	CE1 Ethernet1	100.0.1.1/24	PE1 Ethernet	100.0.1.2/24
PE1-ASBR1	PE1 Ethernet2	1002::1/64	ASBR1 Ethernet2	1002::2/64
ASBR1-P1	ASBR1 Ethernet3	1003::1/64	P1 Ethernet1	1003::2/64
ASBR1-P2	ASBR1 Ethernet4	1010::2/64	P2 Ethernet1	1010::1/64
ASBR2-P1	ASBR2 Ethernet3	2003::2/64	P1 Ethernet2	2003::1/64
ASBR2-P2	ASBR2 Ethernet4	2010::2/64	P2 Ethernet2	2010::1/64
ASBR2-PE2	ASBR2 Ethernet2	2002::1/64	PE2 Ethernet2	2002::2/64
PE2-CE2	PE2 Ethernet1	200.0.0.2/24	CE2 Ethernet1	200.0.0.1/24

Loopback Address and Locator Prefix Plan

Device	Loopback0	Locator Prefix	SID	Behavior
ASBR1	2000::1/64	2000::/64	2000::f011:0:0:0/80	End
P1	2000:0:0:1::/64	2000:0:0:1::/64	2000:0:0:1:e001::/80	End
P2	2000:0:0:2::/64	2000:0:0:2::/64	2000:0:0:2:e001::/80	End
ASBR2	2000:0:0:3::/64	2000:0:0:3::/64	2000:0:0:3:fe00::/80	End
PE1	2000:0:0:4::/64	2000:0:0:4::/64	2000:0:0:4:e002::/80	End.DT4
PE1	2000:0:0:4::/64	2000:0:0:4::/64	2000:0:0:4:e000::/80	End
PE2	2000:0:0:5::/64	2000:0:0:5::/64	2000:0:0:5:e003::/80	End.DT4
PE2	2000:0:0:5::/64	2000:0:0:5::/64	2000:0:0:5:e000::/80	End

Configuration Roadmap

Enable IPv6 forwarding on all interfaces of the PE, ASBR, and P devices, and configure interface IPv6 addresses
Enable IS-IS within each AS and configure the level
Configure VRFs on PE1 and PE2
Establish EBGP peer relationships between ASBRs and P devices
Establish MP-BGP peer relationships between PE1 and ASBR1, ASBR1 and ASBR2, and ASBR2 and PE2 to advertise VPN routes
Enable SRv6 on the PE, ASBR, and P devices, and configure Locator prefixes and SIDs
Configure route-maps to assign different Colors to different subnets
Create SR-TE Policies matching the Colors

Service Configuration

Configure Interface and Loopback Addresses

Enable IPv6 forwarding on each interface and configure IPv6 addresses. PE1 is used as an example; the configuration on other devices is similar and is not repeated here.

interface loopback 0
 ip address 2000:0:0:4::1/64
exit

interface ethernet 2
 router-interface
 ip address 1002::1/64
 ipv6 use-link-local
exit

Configure IS-IS

Both AS 65001 and AS 65002 use IS-IS Level-1 internally to build the IPv6 underlay. PE1 and ASBR1 are used as examples; the configuration on PE2 and ASBR2 is similar and is not repeated here.

# Configure PE1

router isis test
 is-type level-1
 net 10.0000.0000.0000.2222.00
 redistribute ipv4 bgp level-1
 redistribute ipv4 connected level-1
 redistribute ipv6 bgp level-1
 redistribute ipv6 connected level-1
exit

interface ethernet 2
 ip router isis test
 ipv6 router isis test
 isis metric level-1 2
 isis network point-to-point
 no isis hello padding
exit

# Configure ASBR1

router isis test
 is-type level-1
 net 10.0000.0000.0000.3333.00
 redistribute ipv4 bgp level-1
 redistribute ipv4 connected level-1
 redistribute ipv6 bgp level-1
 redistribute ipv6 connected level-1
exit

interface ethernet 2
 ip router isis test
 ipv6 router isis test
 isis network point-to-point
 no isis hello padding
exit

After the configuration is complete, verify that IS-IS is configured successfully as follows.

# Display IS-IS neighbor information (PE1 as an example).

show isis neighbor
Area test
System Id           Interface     L     State       Holdtime SNPA
 ASBR1                eth 2  1    up              29       2230.6602.375c

Configure VRFs and CE Access

Create VRF 100 on both PEs and establish an IPv4 IBGP neighbor with the CE within the VRF. CE1 and PE1 are used as examples below; the configuration on CE2 and PE2 is similar.

# Configure CE1

interface ethernet 1
router-interface
ip address 100.0.1.1/24
exit

interface ethernet 3
router-interface
ip address 10.0.0.2/24
exit

router bgp 65001
no bgp default ipv4-unicast
no bgp ebgp-requires-policy
neighbor 100.0.1.2 remote-as internal
address-family ipv4 unicast
neighbor 100.0.1.2 activate
redistribute connected
exit
exit

# Configure PE1

vrf 100
exit

interface ethernet 1
router-interface
ip address 100.0.1.2/24
vrf 100
exit

router bgp 65001 vrf 100
bgp router-id 1.1.1.1
no bgp default ipv4-unicast
no bgp ebgp-requires-policy
neighbor 100.0.1.1 remote-as internal
address-family ipv4 unicast
export vpn
import vpn
neighbor 100.0.1.1 activate
rd vpn export 3:20
rt vpn both 88:88
exit
exit

Configure EBGP Peers Between ASBRs and P Devices

Configure EBGP peers between ASBRs and P devices. ASBR1 and P1 are used as examples; the configuration between ASBR1 and P2, P1 and ASBR2, and P2 and ASBR2 is similar.

# Configure ASBR1

router bgp 65001
 no bgp default ipv4-unicast
 no bgp ebgp-requires-policy
 neighbor 1003::2 remote-as 65000
 neighbor 1003::2 capability extended-nexthop
 neighbor 1003::2 ebgp-multihop 10
 neighbor 1003::2 update-source 1003::1
 address-family ipv4 unicast
  neighbor 1003::2 activate
  redistribute connected
 exit
 address-family ipv6 unicast
  neighbor 1003::2 activate
  redistribute connected
  redistribute isis
 exit
exit

# Configure P1

router bgp 65000
 no bgp default ipv4-unicast
 no bgp ebgp-requires-policy
 egress-engineering enable
 neighbor 1003::1 remote-as 65001
 neighbor 1003::1 capability extended-nexthop
 neighbor 1003::1 ebgp-multihop 10
 neighbor 1003::1 update-source 1003::2
 address-family ipv4 unicast
  neighbor 1003::1 activate
  redistribute connected
 exit
 address-family ipv6 unicast
  neighbor 1003::1 activate
  redistribute connected
 exit
exit

Configure MP-EBGP/MP-IBGP Peers

Using PE1 and ASBR1 as an example, configure MP-IBGP peers to advertise VPN routes within the AS. The configuration between PE2 and ASBR2 is similar.

# Configure PE1

router bgp 65001
 bgp router-id 1.1.1.1
 no bgp default ipv4-unicast
 no bgp ebgp-requires-policy
 neighbor 2000::1 remote-as 65001
 neighbor 2000::1 capability extended-nexthop
 neighbor 2000::1 update-source 2000:0:0:4::1
 address-family ipv4 unicast
  redistribute connected
 exit
 address-family ipv4 vpn
  neighbor 2000::1 activate
 exit
 address-family ipv6 unicast
  redistribute connected
  redistribute isis
 exit
exit

# Configure ASBR1

router bgp 65001
 neighbor 2000:0:0:4::1 remote-as 65001
 neighbor 2000:0:0:4::1 capability extended-nexthop
 neighbor 2000:0:0:4::1 update-source 2000::1
 address-family ipv4 vpn
  neighbor 2000:0:0:4::1 activate
  neighbor 2000:0:0:4::1 next-hop-self
 exit
exit

Configure the MP-EBGP peer between ASBR1 and ASBR2 to pass VPN routes across ASs.

# Configure ASBR1

router bgp 65001
 neighbor 2000:0:0:3::1 remote-as 65002
 neighbor 2000:0:0:3::1 capability extended-nexthop
 neighbor 2000:0:0:3::1 ebgp-multihop 10
 neighbor 2000:0:0:3::1 update-source 2000::1
 address-family ipv4 vpn
  neighbor 2000:0:0:3::1 activate
  neighbor 2000:0:0:3::1 next-hop-self
 exit
exit

# Configure ASBR2

router bgp 65002
 neighbor 2000::1 remote-as 65001
 neighbor 2000::1 capability extended-nexthop
 neighbor 2000::1 ebgp-multihop 10
 neighbor 2000::1 update-source 2000:0:0:3::1
 address-family ipv4 vpn
  neighbor 2000::1 activate
  neighbor 2000::1 next-hop-self
 exit
exit

Run show ipv6 bgp summary to check the BGP neighbor status of the device.

# Check the IPv6 BGP neighbor status on ASBR1.

show ip bgp ipv6 summary

IPv6 Unicast Summary:
BGP router identifier 240.127.1.1, local AS number 65001 VRF default vrf-id 0
BGP table version 50 RIB entries 15, using 1920 bytes of memory Peers 2, using 48 KiB of memory

Neighbor  V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd PfxSnt Desc
1003::2   4 65000 3267   3287     50     0   0   01:12:09   5    8           N/A
1010::1   4 65000 1716   1729     50     0   0   1d04h03m   5    8           N/A

Total number of neighbors 2

IPv6 VPN Summary:
BGP router identifier 240.127.1.1, local AS number 65002 VRF default vrf-id 0
BGP table version 0 RIB entries 3, using 384 bytes of memory Peers 2, using 48 KiB of memory

Neighbor       V AS   MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd PfxSnt Desc
2000:0:0:3::1 4 65002 3227    3268     0      0   0   01:13:52     0          0     N/A
2000:0:0:4::1 4 65001 17       19      36     0   0   00:04:21     0           0     N/A

Configure SRv6 Capabilities

Using PE1 as an example, configure the SRv6 Locator and set the encapsulation format to gsid-f4832.

After the Locator prefix is configured, SIDs are generated automatically. SIDs can also be configured statically using the sid sid locator locator-name behavior {End/End.X/End.DT4/End.DT6} command. The configuration on PE2, ASBR1, ASBR2, P1, and P2 is similar:

# Configure PE1

segment-routing
 srv6
  locators
   locator pe1
    format gsid-f4832
    prefix 2000:0:0:4::/64 block-len 48 node-len 16 arg-len 48
   exit
  exit
 exit
exit

Enable SRv6 in the IS-IS process and associate the Locator. The configuration on PE2, ASBR1, and ASBR2 is similar:

# Configure PE1

router isis test
 segment-routing srv6
  locator pe1
 exit
exit

Enable SRv6 in the BGP process and associate the local Locator ID so that BGP uses this Locator for SRv6 VPN services. The configuration on PE2, ASBR1, and ASBR2 is similar:

# Configure PE1

router bgp 65001
 segment-routing srv6
  locator pe1
 exit
exit

router bgp 65001 vrf 100
 address-family ipv4 unicast
  sid vpn export auto
 exit
exit

If SIDs are allocated dynamically, run the show segment-routing srv6 local sid command after the configuration to check SID allocation.

Configure Route-map

Taking the route for subnet 20.0.0.0/24 advertised from PE2 to PE1 as an example, create a route-map that sets the color of routes matching this subnet to 100.

ip prefix-list test seq 10 permit 20.0.0.0/24
route-map test permit 10
 match ip address prefix-list test
 set extcommunity color 00:100
exit
router bgp 65002 vrf 100
 address-family ipv4 unicast
  route-map vpn export test
  exit
 exit

After the configuration is complete, verify that the VPN routes carry the specified Color as follows.

# View VPN route information on ASBR2.

show bgp ipv4 vpn
BGP table version is 28, local router ID is 6.6.6.6, vrf id 0
Default local pref 100, local AS 65002
Status codes:  s suppressed, d damped, h history, u unsorted, * valid, > best, = multipath,
               i internal, r RIB-failure, S Stale, R Removed
Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
Origin codes:  i - IGP, e - EGP, ? - incomplete
RPKI validation codes: V valid, I invalid, N Not found

     Network          Next Hop            Metric LocPrf Weight Path
Route Distinguisher: 3:20
 *>i 10.0.0.0/24      2000:0:0:3::1                 100      0 65001 ?
    UN=2000:0:0:3::1 EC{88:88} label=917536 sid=2000:0:0:4:: sid_structure=[48,16,16,48] type=bgp, subtype=0
 *>i 10.1.0.1/32      2000:0:0:3::1                 100      0 65001 ?
    UN=2000:0:0:3::1 EC{88:88} label=917536 sid=2000:0:0:4:: sid_structure=[48,16,16,48] type=bgp, subtype=0
 *>  20.0.0.0/24      200.0.1.2@15<            0    100      0 ?
    UN=200.0.1.2 EC{88:88 Color:00:100} label=917552 sid=2000:0:0:5:: sid_structure=[48,16,16,48] type=bgp, subtype=5
 *>i 100.0.1.0/24     2000:0:0:3::1                 100      0 65001 ?
UN=2000:0:0:3::1 EC{88:88} label=917536 sid=2000:0:0:4:: sid_structure=[48,16,16,48] type=bgp, subtype=0

Displayed 4 routes and 4 total paths

Configure TE-Policy

Taking services for subnet 20.0.0.0/24 as an example, create a TE-Policy that specifies the forwarding path for this subnet as: PE1 -> ASBR1 -> P1 -> ASBR2 -> PE2.

# Configure the TE-policy for color 100 on PE1.

segment-routing
 traffic-eng
  segment-list testpath
   index 3 ipv6-address 2000::f011:0:0:0 compress csid-len 32 block-len 48
   index 5 ipv6-address 2000:0:0:1:e001:: compress csid-len 32 block-len 48
   index 10 ipv6-address 2000:0:0:3:fe00:: compress csid-len 32 block-len 48
   index 15 ipv6-address 2000:0:0:5:e001:: compress csid-len 32 block-len 48
  exit
  policy test color 100 endpoint 2000::1
   candidate-path 100 preference 100 explicit segment-list 100 testpath
  exit
 exit

After the configuration is complete, check the route information as follows.

# View route information on PE1.

show ip route vrf all
Codes: K - kernel route, C - connected, L - local, S - static,
       R - RIP, O - OSPF, I - IS-IS, B - BGP, E - EIGRP, N - NHRP,
       T - Table, v - VNC, V - VNC-Direct, A - Babel, F - PBR,
       f - OpenFabric, t - Table-Direct,
       > - selected route, * - FIB route, q - queued, r - rejected, b - backup
       t - trapped, o - offload failure

IPv4 unicast VRF 100:
B>* 10.0.0.0/24 [200/0] via 100.0.1.1, eth 1, weight 1, 01:54:02
B>* 10.1.0.1/32 [200/0] via 100.0.1.1, eth 1, weight 1, 01:54:02
B>  20.0.0.0/24 [200/0] via 2000::1 (vrf default) (recursive), SR-TE color 100, seg6 2000::f011:0:0:0,2000:0:0:1:e001::,2000:0:0:3:fe00::,2000:0:0:5:e001::,2000:0:0:5:e003::, weight 1, 00:08:57
  *                       via fe80::2030:66ff:fe02:375c, eth 2 (vrf default), seg6 2000::f011:0:0:0,5:e003:5:e001:3:fe00:1:e001, encap behavior H.Encaps.Red, weight 1, 00:08:57
B   100.0.1.0/24 [200/0] via 100.0.1.1 inactive, weight 1, 01:25:15
C>* 100.0.1.0/24 is directly connected, eth 1, weight 1, 01:54:06

Verification

Ping PC2 (20.0.0.2) on the right from PC1 to verify inter-AS connectivity for Color 100 services.

PC1> ping -a 20.0.0.2

84 bytes from 20.0.0.2 icmp_seq=1 ttl=60 time=39 ms
84 bytes from 20.0.0.2 icmp_seq=2 ttl=60 time=27 ms
84 bytes from 20.0.0.2 icmp_seq=3 ttl=60 time=47 ms
84 bytes from 20.0.0.2 icmp_seq=4 ttl=60 time=27 ms
84 bytes from 20.0.0.2 icmp_seq=5 ttl=60 time=37 ms
84 bytes from 20.0.0.2 icmp_seq=6 ttl=60 time=35 ms

During the ping test, capture service packets between PE1 and ASBR1 and between ASBR1 and P1 to confirm G-SID encapsulation.

# Packets captured at the PE1 egress:

# Packets captured at the ASBR1 egress: