- Newest
- Most votes
- Most comments
You should set your on-premises routers explicitly to advertise only 0.0.0.0/0 over BGP towards AWS, instead of advertising all the hundreds of routes that the routers know about natively.
Within your VPC, you can have a static 0.0.0.0/0 route pointing to your virtual private gateway (for which the standard abbreviation is VGW, not VPG) or let the on-premises-advertised default route propagate to your VPC's route tables. It doesn't make a functional difference which of these options you choose, as long as there are no competing routes to take priority over.
You should place a direct connect gateway (DXGW) between your direct connect VIF and the VGW, instead of connecting the VIF directly to the VGW. The latter will technically work, but the former is recommended, because it allows AWS's network backbone to identify multiple connections associated with the same DXGW as being meant to provide redundancy. Plugging multiple VIFs directly to a VGW isn't guaranteed to do so.
In the routing tables for the VPC/Subnets, you can create a default route of 0.0.0.0/0 to the VPG.
Its very straight forward
Hi Gary, This is already defined but any destination outside learned BGP routes and Site-Site VPNs is not going through AWS Direct connect connection with destination unreachable error locally itself. Devices within VPC are reaching to VPG for any destination because of the default route but getting dropped then after.
You'll need BGP between on premises and AWS @Boinpally. The static route from the VPC to the VGW won't take the packets any further than the VGW.
Appologies I miss understood.. Thanks Leo, your correct
Relevant content
- Accepted Answerasked 3 years ago
- Accepted Answer
asked 6 months ago - Accepted Answerasked a year ago
AWS OFFICIALUpdated a year ago- AWS OFFICIALUpdated 2 years ago
- AWS OFFICIALUpdated 2 months ago
Hi Leo, Thank you.
I have challenge of originating a default route from other end. As I can't rely on BGP due to routes limitation on Private VIF, having a static default route defined to AWS Hosted direct connect connection would help.
Instead of VGW, Transit VG would help in this regard? I think this needs re-association through Transit VIF from my partner but can define a static route but not sure whether I can point static default route to AWS Hosted direct connect connection?
Hi Boinpally, I'm afraid there's no option but to use BGP between on-prem and AWS. Also with transit gateways (TGW), a DXGW between it and on-premises would have to exchange routes over BGP. In my professional experience, advertising the default route 0.0.0.0/0 or broadly summarised internal networks, such as 10.0.0.0/8, has worked without any issues for closer to 10 years, and the design avoids the issue of having to deal with the limitations on the number of advertised/received routes.
Hi Leo, Thank you. Let me put further detail here.
AWS India (VPC) <-> AWS DX Private VIF <-> Partner MPLS <-> Our MPLS <-> AWS DX Public VIF <-> AWS Ireland/Genesys EU The objective is to receive Genesys Public subnets (right hand side) into our VPC in AWS India VPC. We tested with only 3 routes at our partner proving Hosted connection to us and now think that Genesys BGP route getting dropped bcoz of AWS ASN for 52.129.106.0/24 after receiving at our VPC in India as we see other 2 x routes successfully. show router xx bgp neighbor x.x.x.105 advertised-routes
BGP IPv4 Routes
Flag Network LocalPref MED Nexthop (Router) Path-Id Label As-Path
i 52.129.106.0/24 n/a None 10.1.1.101 None - xx xx xx 7224 7224 16509 i 10.1.1.0/24 n/a None 10.1.1.101 None - xx 65100 i 10.5.50.0/24 n/a None 10.1.1.101 None - xx 65100
Routes : 3
Any inputs? Ashok.