> For the complete documentation index, see [llms.txt](https://ccie-sp.gitbook.io/ccie-spv5.1-labs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ccie-sp.gitbook.io/ccie-spv5.1-labs/labs/bgp/pic-edge-for-vpnv4.md).

# PIC Edge for VPNv4

**Topology**: ine-spv4

<figure><img src="/files/W7z2NmEvDwMO2ibFZjLQ" alt=""><figcaption></figcaption></figure>

Load **bgp.pic.init.cfg**

```
#IOS-XE
config replace flash:bgp.pic.init.cfg
 
#IOS-XR
configure
load bootflash:bgp.pic.init.cfg
commit replace
y
```

VPNv4 is currently setup in the network. R2 is a dual-homed CE that announces 1.1.1.1/32 and 2.2.2.2/32 to R3 and R4. XR1 is singled homed to R5, announcing 19.19.19.19/32 and 20.20.20.20/32. R6 is the RR.

Configure PIC for the L3VPN so that R5 pre-installs a backup route in the FIB for 1.1.1.1/32 and 2.2.2.2/32 for VRF A.

## Answer <a href="#bdada2b4-5839-49ac-992e-d2a80b25389d" id="bdada2b4-5839-49ac-992e-d2a80b25389d"></a>

```
#R3
vrf def A
 no rd 100:1
!
! Wait ~30 seconds
!
 rd 100:3
 route-target both 100:1
!
router bgp 100
 add ipv4 uni vrf A
  neighbor 20.2.3.2 remote-as 65000

#R5
router bgp 100
 add ipv4 uni vrf A
  bgp additional-paths install
```

## Explanation <a href="#ac1804d6-5003-4493-b24d-a00d2949bc71" id="ac1804d6-5003-4493-b24d-a00d2949bc71"></a>

In this lab, we must use per-PE VRF RD values to get R6 to reflect the multiple paths to R5. IOS-XE does not support sending additional paths for VPNv4. Only IOS-XR supports this. So if IOS-XR was the RR, and the PE we want to run PIC was also IOS-XR, we could have used the add-path capability for this instead.

However, VPNv4 gives us an easy way to make prefixes unique, which is to use a different RD per PE. When you change the RD value, you must wait a while for the router to deconfigure the VRF. Then when it is done, you must put in the new RD value, re-add the RTs, and reconfigure the CE as a neighbor under BGP.

```
#R3
vrf def A
 no rd 100:1
!
! Wait ~30 seconds
!
 rd 100:3
 route-target both 100:1
!
router bgp 100
 add ipv4 uni vrf A
  neighbor 20.2.3.2 remote-as 65000
```

At this point R5 should have two paths to 1.1.1.1/32 and 2.2.2.2/32. All that is left is to tell R5 to install additional paths as backup routes using the command **bgp additional-paths install** under the customer VRF under BGP.

```
#R5
router bgp 100
 add ipv4 uni vrf A
  bgp additional-paths install
```

## Verification <a href="#c1a5f59e-faef-44e3-a87e-741ebf2187c0" id="c1a5f59e-faef-44e3-a87e-741ebf2187c0"></a>

Before we make any changes, we can see that R5 only has one route to prefixes advertised by R2.

<div align="left"><figure><img src="/files/J4uegwYPt1zHlgXzfwhS" alt=""><figcaption></figcaption></figure></div>

This is because the RR will only select the bestpath to reflect to clients:

<div align="left"><figure><img src="/files/wq7vi6Y8dnBWWYNcgKDo" alt=""><figcaption></figcaption></figure></div>

When we change the RD on R3, the RR sees this as a completely separate prefix due to the unique RD value. Therefore it reflects both paths to its clients.

<div align="left"><figure><img src="/files/0bvYU9LD4mv3wdyPBvAe" alt=""><figcaption></figcaption></figure></div>

On R5, we now have multiple paths, but we still have not enabled PIC yet.

<div align="left"><figure><img src="/files/APFPyxJ064jN9m0sxa8o" alt=""><figcaption></figcaption></figure></div>

We enable PIC for a VRF under the BGP VRF config:

```
#R5
router bgp 100
 add ipv4 uni vrf A
  bgp additional-paths install
```

Once PIC is enabled, we should see a **b** next to the backup paths:

<div align="left"><figure><img src="/files/OPaKT0cDtytSvN1WHbrG" alt=""><figcaption></figcaption></figure></div>

We also see the backup/repair path in the FIB:

<div align="left"><figure><img src="/files/vtPYoU0yHYLC5TxlASvn" alt=""><figcaption></figcaption></figure></div>

This can also be seen in the RIB using the following command:

<div align="left"><figure><img src="/files/va14ezkxbuNxVairB0wC" alt=""><figcaption></figcaption></figure></div>

## Further Reading <a href="#id-9dcc8341-694f-4115-9446-8a0d01dfbc25" id="id-9dcc8341-694f-4115-9446-8a0d01dfbc25"></a>

<https://learningnetwork.cisco.com/s/question/0D53i00000Kt0rfCAB/multipath-in-bgp-vpnv4-network>
