Hi Aijun,
> Along the discussions of this document(and also its LSR companion document > (https://datatracker.ietf.org/doc/draft-many-lsr-power-group/), is there > anyone consider the possibilities of power-aware TE in distributed manner > plausible? Why yes, I consider it plausible and even likely, given that it works. > In the document, it states: > "PCPPS leverages Constraint-based Shortest Path Forwarding (CSPF). > CSPF can be centralized or distributed. When it is centralized, a > Path Computation Element (PCE) calculates a path for every flow in > the network. When it is distributed, each node calculates a path for > each flow that originates on it." > > But, in distributed mode, if "each node calculate a path for each flow that > originates on it", how can this node know the traffic from other nodes do not > pass one of the sleep-capable interfaces that it just moves out of the > traffic? That information is not necessary. Each node is free to make its own decisions. Assuming that the head ends are able to shift the traffic off of the power-sleep-capable link and onto other available capacity, the relevant nodes can then detect that the link is idle and can put it into power-sleep. > If the nodes within the network can't coordinate their decisions, I think it > is impossible to achieve the distributed power-aware TE aim. Distributed independent decisions need not be explicitly coordinated to have the desired effect. If you are in the middle of a flock of birds and clap your hands, they will fly away. > Please also be reminded, RSVP style TE is fading out of the network due to > its stateful table stress. Patently false. Our customer base continues to grow nicely, thank you very much. > Such aim can only achieved via the centralized mode, or via the management > system. Again, provably false. We have working code. Regards, Tony _______________________________________________ Lsr mailing list -- [email protected] To unsubscribe send an email to [email protected]
