BGP Auto-peering Feature Description
Note
BGP auto-peering relies on a default STPD instance "s0" to resolve
loops on the default VLAN.
Note
Auto-peering automatically creates VLANs
dynamically when using the
create auto-peering bgp
routerid
ipaddress
AS-number
asNumber
command.
Note
The command
create auto-peering bgp
routerid
ipaddress
AS-number
asNumber
uses
eBGP adjacencies. The AS number should be different for all devices participating in the
network.
After executing the command create auto-peering bgp
routerid
ipaddress
AS-number
asNumber
, the
following series of events occurs, setting up the auto-peering:
- AutoBGP capability is advertised out all ports by LLDP automatically. AutoBGP gets its
configuration using cloud connector through Zero Touch Provisioning. A Python script initiates
default VLAN and BGP configuration that is needed for AutoBGP.
- When an AutoBGP-capable neighbor is detected out a port, LLDP notifies the AutoBGP
manager.
- The AutoBGP manager assigns the LLDP-learned port to one of the VLANs within its configured
range and notifies VLAN manager. DHCP Relay is enabled on this VLAN.
- AutoBGP informs LLDP of the IPv6 link-local address configured on that VLAN along with the
BGP router ID and AS number.
- LLDP sends the AutoBGP information (IPv6 link-local, router ID, and AS number ) to the remote
neighbor and learns the remote neighbor information and then send this to the AutoBGP
manager.
- The AutoBGP manager creates an AutoBGP route to the learned remote router ID in route
manager.
- The AutoBGP manager creates and enables the BGP neighbor.
- The AutoBGP is formed and host route updates occur allowing end-to-end connectivity.
- MBGP capability to carry VXLAN information is automatically enabled on each BGP peer when
AutoBGP is enabled. In addition BGP is registered as a client of the Overlay Tunnel Manager
(OTM), which maintains VNI/LTEP information.
- When OTM learns VNI/LTEP pairs, these are passed to the BGP client, which generates BGP
advertisements to any peers that the MBGP/VXLAN capability has negotiated.
- When a BGP update is received containing a VNI/LTEP pair, the BGP client passes this to OTM
using the client interface.
- OTM drives IBGP session to remote VTEP leaf to support EVPN required for VXLAN tunnel.