Hi Cyril,

Thanks a lot for your feedback!

Maybe I need to make it clear that the problem we’re trying to solve is a 
single optimization objective resulting in multiple ECMP/UCMP paths. This is 
motivated by SR-TE Policy use case, where each Candidate Path represents a 
single optimization objective. The Candidate Path has a set of Segment Lists 
that satisfy the optimization objective.

You seem to want to solve a different problem: two or more different 
optimization objectives and each ECMP path is mapped to a different objective. 
In that case Solution 1 is absolutely necessary and it would not have any of 
the down-sides, because the PCC knows in advance how many optimization 
objectives it has and can create that many PCEP LSPs. However for our problem, 
Solution 1 would introduce a lot of implementation complexity and protocol 
overhead.

We have a side-meeting scheduled on Wednesday at 8:30 to discuss this topic, 
you are welcome to attend if you want to contribute your input.

Thanks,
Mike.

From: Cyril Margaria <[email protected]>
Sent: Friday, July 19, 2019 9:37 AM
To: Mike Koldychev (mkoldych) <[email protected]>
Cc: [email protected]
Subject: Re: [Pce] Proposal for signaling ECMP or UCMP paths

Hi,

On slide "LSP objectives and constraints": Stateless  PCE can compute set of 
EROs/Label switch paths using RFC6007, including multi-domain and multi-PCEs 
scenarios. This can be used for computing a set of EROs for SR candidate paths, 
one case that can apply to the candidate path and explicitly mentioned by the 
RFC is "Two or more end-to-end diverse paths".  This does not cover the 
stateful PCE case directly, but there are similar situations to what RFC6007 in 
the form of path protection (primary/secondary/standby) for statefull PCE, 
which use the association mechanism. Those two existing mechanism exists to 
coordinate several paths and could be used to indicate how multiple paths are 
related and on how to signal them together (SVEC)

On slide "Analysis of Solution 1":
  - For PCC-Initated LSPs: what prevents the PCE to to create PCE-Initiated 
LSPs using the same association id? This would tackle the problem.
  - The possibility of each path to have different objective does seems to be 
an advantage as its less restrictive. Having the same restriction on a set of 
paths is easy, relaxing a restriction on the ERO #5 is more complicated (in 
term of encoding).
  - There is a set of options to achieve the "signal the set of paths together":
     a)  set of LSPs can be reported in the same message, it can be enforced by 
the document defining that specific association type.
     b) SVEC/SVEC List can be extended to statefull PCEP,

That solution would work in case of multi-domain PCEs, and also be helpful for 
OAM and auto-bw mechanism.
As a segment list is one path in the network, that maps nicely to one LSP.


Solution 2:
  - This limit the set of constraints to be applied, policies like "10% of the 
traffic does not need to be protected" cannot be expressed (it can be with 
solution 1, clear L bit of LSPA on one TE-LSP out of 10)
  - 2.a when the LSP is reported down : which ERO is down?, the same is 
applicable for auto-bw and any form of OAM data.
  - Solution 2.b allows for Optimized branch encoding, that should be disabled 
for that solution


Slide "Comparison of Solutions":
   - There are solutions to most of the points raised for solution 1
   - The database problem seems specific to one implementation, other 
implementation will have the problem for solution 2
   -  multi-PCE and multi-domain are not evaluated. Solutions and consideration 
are available for solution 1, not for solution 2. (experimental Inter-domain 
P2MP tree solutions exists).

Best Regards,
Cyril

On Fri, 12 Jul 2019 at 22:02, Mike Koldychev (mkoldych) 
<[email protected]<mailto:[email protected]>> wrote:
Hi WG,

As per SPRING WG, SR Policy may contain multiple Candidate Paths and each 
Candidate Path may contain multiple Segment Lists. Existing SR standards in 
PCEP allow only a single ERO (one Segment List) for the SR Path in a stateful 
PCEP message. There is a need to signal multiple Segment Lists in PCEP for this 
as well as other load balancing use cases.

See the link that describes this, as well as list possible ways to achieve 
this. Please provide your feedback on the list or during the WG session.

https://github.com/dhruvdhody-huawei/105/blob/master/multiple_ERO_jl03a.pdf

Thanks,
Mike.

_______________________________________________
Pce mailing list
[email protected]<mailto:[email protected]>
https://www.ietf.org/mailman/listinfo/pce
_______________________________________________
Pce mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/pce

Reply via email to