Hi Adrian, > On Feb 13, 2021, at 12:34 PM, Adrian Farrel <[email protected]> wrote: > > I have some pretty strong opinions on the idea of a single node > abstraction. The main challenge comes when there is a partial failure in > the zone such that the zone is partitioned (or the path between two > zone neighbors across the zone is severely degraded). It is not possible > to represent this in the node model since your only options are: > - drop the connection to a neighbor > - move to represent the zone as two nodes
I’m not one of the authors and I don’t mean to interfere, but I have to ask whether you’re raising the bar here. Partitioned areas have been a sore spot within IS-IS since day one. While we’ve frequently discussed alternative solutions (e.g., tunneling through L2 to restore L1 continuity), I’m unaware of a practical, deployed solution to the problem. We have always addressed this by placing the burden on network architects: ensure that your L1 area has enough redundancy to never partition. While this seems like it is annoying, it does seem to be doable. While it is certainly fair to call this out as a discussion item, requiring that TTZ (and Area Proxy) solve issues that we haven’t been able to solve to date seems to be asking for technological breakthroughs as table stakes. Regards, Tony
_______________________________________________ Lsr mailing list [email protected] https://www.ietf.org/mailman/listinfo/lsr
