Configuring BGP Advanced Settings
This topic describes how to manage advanced Border Gateway Protocol (BGP) settings in the Megaport Cloud Router (MCR).
- The Multi-Exit Discriminator (MED) path attribute to identify the preferred VXC used to re-route traffic after a path failure.
- The Bidirectional Forwarding Detection (BFD) protocol to detect any path failures between BGP neighbors.
Configuring the BFD protocol
BGP shares network reachability information with adjacent BGP systems, which are referred to as neighbors or peers. The BFD protocol detects path failure between directly connected BGP neighbors, allowing for a faster BGP routing reconvergence time. Enabling BFD on a VXC connection provides fast link failure detection and failover when connecting to services that support BFD on the remote neighbor. After BFD is enabled, a BGP neighbor relationship can be torn down quickly after notifications from BFD, failing over to another BGP neighbor.
The BFD protocol runs independent of BGP. When BFD is enabled, failover is faster because it occurs before the BGP timer interval elapses. The default setting for the BGP timer is 3 keep-alive window failures of 60 seconds each. When not enabled, failover will only occur after the BGP timer interval elapses.
BFD settings apply to all BGP connections on the interface.
Important
To prevent high CPU utilization, before you enable BFD, disable the sending of Internet Control Message Protocol (ICMP) redirect messages by entering the no ip redirect
command.
To enable BFD
-
Select the VXC and select the A-End or B-End Configuration tab.
-
Click Edit.
-
Next to the BGP connection, click the gear icon
and select Edit.
-
Select the Advanced tab.
-
Next to Enable BFD, click Enabled.
Enabling BFD configures the default settings: the TX interval is set to 300 milliseconds, the RX interval is set to 300 milliseconds, and the Multiplier is set to 3. You can adjust these settings later.
-
Click Confirm.
Important
You must enable BFD on the BGP neighbor as well.
Note
BFD must be enabled to modify BFD settings.
To modify the BFD settings
-
Select the VXC and select the A-End or B-End Configuration.
-
Click Edit.
-
After TX Interval, specify the minimum time that the BGP neighbor transmits liveness detection BFD packets to the BGP neighbor. The default is 300 milliseconds. The range is 300 milliseconds to 9000 milliseconds.
-
After RX Interval, specify the minimum time that the BGP neighbor sends liveness detection BFD packets to this interface. The default is 300 milliseconds. The range is 300 milliseconds to 9000 milliseconds.
-
After Multiplier, specify the minimum number of BFD packets that can be missed before the BGP session is considered down. The default is 3 packets. The range is 3 to 50 packets.
-
Click Confirm.
Default BFD values by cloud service provider
-
AWS Direct Connect – BFD is supported natively by AWS Direct Connect connections. Asynchronous BFD is automatically enabled for AWS Direct Connect virtual interfaces on the CSP side. However, you must configure BFD on MCR VXCs to AWS Direct Connect to enable it for your connection.
By default, the AWS BFD liveness detection minimum interval is 300 milliseconds and the multiplier is 3 packets.
Note
Confirm support for other neighbors with your CSP, as well as specific default values.
-
Azure ExpressRoute – BFD is supported natively by Azure ExpressRoute on private peering. BFD is configured by default under all the newly created ExpressRoute private peering interfaces on the MSEEs. However, you must configure MCR VXCs to Azure ExpressRoute private peering for BFD to enable it for your connection. Between BFD peers, the slowest peer determines the transmission rate.
-
Microsoft Enterprise Edge (MSEE) ExpressRoute routers – BFD transmit and receive intervals are set to 300 milliseconds. In certain scenarios, you can set the intervals to a higher value of 750 milliseconds. You can set higher values to increase the intervals but you can’t configure values lower than 300 milliseconds to decrease them.
-
Google Cloud Services – Google is testing BFD in certain locations, but not in a manner Megaport can support until they offer it everywhere. Megaport will monitor and add support as soon as practical.
Configuring a preferred route with MED
A Multi-Exit Discriminator (MED) is a BGP path attribute that can influence a BGP neighbor to take a preferred route when the advertising autonomous system (AS) is the same for candidate routes and there are multiple entry points for that AS. A lower MED metric is preferred over a higher metric. You can use the MED attribute to switch traffic between two VXCs.
Other BGP attributes are taken into account before considering the MED attribute. The MED attribute breaks the tie between two routes when the weight, local preference, AS path, and origin type are the same. The exit point with the lowest MED metric is preferred.
For example, a configuration that consists of a BGP neighbor device, R3, with a MED metric of 20 and another neighbor device, R2, with a MED metric of 10 will cause autonomous system (AS) 123 (of which device R2 and R3 are members) to prefer the path through R2 to reach ASN4.
Adding or changing an MED metric doesn’t disrupt the service.
To configure the MED
-
Select the VXC and select the A-End or B-End Configuration tab.
-
Click Edit.
-
Next to the BGP connection, click the gear icon
and select Edit.
-
Select the Advanced tab.
-
Next to Inbound, specify a 32-bit integer MED metric to apply to all routes received on this BGP connection. Leave blank to use the value received from the BGP neighbor.
-
Next to Outbound, specify a 32-bit integer MED metric to apply to all routes transmitted by this BGP connection.
-
Click Confirm.
Using MED with Google Cloud Platform
Google Cloud Platform Cloud Router egress is determined by the first condition that is met:
- All egress traffic is sent to the route with the shortest AS path length.
- If routes have the same AS path lengths, all egress traffic is sent to the route with the lowest MED value (the lowest route metric).
- If routes have the same AS path lengths and MED values (the same route metrics), egress traffic is balanced across all routes using equal-cost multipath (ECMP).
Note
Confirm support for other neighbors with your CSP, along with any specific routing metric influences that might have a higher priority than the MED.