Add multicast-focused AFTs#1468
Conversation
|
FWIW, I swear I had opened a similar PR/ticket back in 2018....ish, but I can't find it now. edit: found it |
There was a problem hiding this comment.
Code Review
This pull request introduces support for multicast forwarding tables (AFT) within the OpenConfig model, specifically adding submodules for IPv4, IPv6, and MPLS P2MP multicast forwarding. It defines new groupings for multicast outgoing interface states and structural lists, which are integrated into the main AFT module. Review feedback recommends utilizing existing common groupings in the new IPv4 and IPv6 submodules to ensure consistency and minimize code duplication.
d5ccef2 to
fd48ac6
Compare
|
Pushed a new changeset based on feedback received. Using the existing NHG pattern has the side effect of removing RPF information from the model. This is important information to have (traffic received on the non-RPF interface isn't forwarded, usually punted to CPU/control plane). Where is the right place to put this information? Should it just be a leaf in the mc-entry/state grouping? Somewhere in the AFT-common tree? |
|
Just noticed I need to incorporate some recent changes in the MPLS modeling into this change, working on that now. |
|
Additional question - unicast MPLS AFT doesn't include |
As an operator, I need a standard way to monitor the current forwarding behavior of multicast on my network, across multiple vendor types.
Most(?) platforms forward multicast under a input interface / OIL model, keyed by various FECs. This change attempts to provide an initial implementation for IPv4, IPv6, and MPLS forwarding of multicast traffic.
Change Scope
Platform Implementations
Tree View