RE: RPR protection with MPLS

Definitely.

-----Original Message-----
From: A.Herrera [mailto:[email protected]]
Sent: Monday, November 18, 2002 11:21 PM
To: John Lemon; [email protected]; [email protected]
Subject: Re: [IPORPR] RPR protection with MPLS


John,

I agree, protection kicks in quickly and there's no
need nor time for MPLS to intervene. I was referring
more to other scenarios where multiple faults occur
with multiple Signal Fails or combinations of Signal
Degrades and instances where a few nodes are isolated
creating 'island' sub-rings.

In these cases, the WG might like to look at scenarios
where MPLS might compliment RPR protection.

Albert


----- Original Message -----
From: <[email protected]>
To: <[email protected]>; <[email protected]>; <[email protected]>
Sent: Monday, November 18, 2002 1:51 PM
Subject: RE: [IPORPR] RPR protection with MPLS


Albert,

I'm not sure what you mean by "RPR protection settles". There is no "settling"
for protection. It just switches to the alternate path, as quickly as the link
problem is detected. Since the RPR MAC owns the interface to the physical layer,
the RPR MAC will learn first and react first.

jl

-----Original Message-----
From: A.Herrera [mailto:[email protected]]
Sent: Tuesday, November 12, 2002 4:17 PM
To: nitin panjwani; David Zelig; 'Frank Kastenholz'; [email protected]
Subject: Re: [IPORPR] RPR protection with MPLS


----- Original Message -----
From: "nitin panjwani" <[email protected]>
To: "David Zelig" <[email protected]>; "'Frank Kastenholz'"
<[email protected]>; <[email protected]>
Sent: Tuesday, November 12, 2002 3:09 PM
Subject: RE: [IPORPR] RPR protection with MPLS


>
> Hi David and all,
> Shouldn't it be something like: either RPR or MPLS
> protection will take charge of rerouteing the traffic;
> and one which assures minimum packate loss and delay
> will win and take charge.
> On this, usually upper layer(MPLS) depends on the
> lower layer(RPR)(as it simplifies the implementation
> and serves the perfect goal of layering approach)and
> shouldn't it be upto MPLS to decide wheathet to take
> charge of protection or wait for RPR, depending on how
> critical network and the passing traffic is(this
> should be configurable).
> Please correct me.
>
> Tahnks,
> Nitin
>
Both MPLS FRR and RPR steer/wrap are localized
reactions to divert within ms. to a pre-determined
detour path or alternate ring direction. So I'll
assume that operator preferences to fault scenarios
are pre-determined and pre-configured as per network
planning based on critical flows.

In this case, I would think, that sufficient feedback from
RPR as to the nature of the faults (i.e. multiple or
combinations of Signal Fail/Degrade, etc.) will allow MPLS
to choose to divert even before RPR protection settles.
For single occurrances of SF, etc, RPR is more than
sufficient protection for critical flows.

These mechanisms and scenarios should be discussed
and further clarified within the WG.

Albert


_______________________________________________
IPORPR mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/iporpr