For cascaded inservice upgrades for
routing protocols(like.. OSPF)
running in an ISP.
In order to prevent any
forwarding loss, I would perform a graceful restart.
However, since these
mechanisms are planned but are not deterministic
as to always minimize
Im working on a method to perform the planning of graceful
better and automatic prior to the actual restart signaling.
want to know both from protocol implementors and operators the value
on this kind of idea.
My ROI comments: Please feel free to add to
1) Give operator confidence in performing planned restarts by making
a controller of minimizing the forwarding
2) ISP's able to predict and inform users or other transit users
the unavailability or loss of connectivity (just like
they do in the TV broadcasts
due to solar flares).
configuration of a set of devices to be automatically helping
and giving a
all is well signal before any planned restarts actually
4) Make operators free of multiple tools which are not always
5) Make this mechanism automated as a part of the protocol ,
a different version or same version with backward
Please give me an idea of ROI if any investment is
made in the
development of such a framework for the protocol community and
operator as a whole.
It will push me in the right
Thanks and Regards,