Could you guys send comments and thoughts in this architectural issue?
The positives and negatives of a tiered P-PE-CE architecture versus a P/PE collapsed solution. See attached diagram.(Not commercial :) )
- Routers are much more powerful now and don't need tiered design so P/PE can handle packet processing as well as edge features.
- Separate P and PEs may give you more availability since PEs are the ones most touched by operations. Ps stay more static in terms of changes.
- Depends on the size of the network? Port density?
- Any concern about MPLS-TE FRR in these architectures.
Any advise on how to calculate that story of 5 nines (99.999%) or so.. Any tool or process?
Appreciate any comments.