The MPLS-OPS Archive

Cell Relay Retreat>MPLS-OPS Archive>month:2002-Feb> msg00168



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

Re: Doubt in MPLS fault-tolerance

  • From: Mathew Lodge <mathew@cplane.com>
  • Date: Mon, 25 Feb 2002 18:23:21 -0800
  • Resent-Date: Mon, 25 Feb 2002 21:56:05 -0500
  • To: "ramu" <ramubachala@hotmail.com>, "MPLS FAQ's" <mpls-ops@mplsrc.com>
  • X-Sender: lodge@127.0.0.1

At 05:16 PM 2/22/2002 -0600, ramu wrote:
Considering fault-tolerance in MPLS,
We will always have back-up path for the current path.
If current path fails, the traffic is shifted onto the backup one.
Once restored the traffic is shifted back to the current one.
Now, this way the load oscillations will occur which is undesirable if =
the failure occurs frequently on the current one, since the probability =
will be higher to failure.

Not sure what you mean by "probability will be higher to failure", but if the backup path is traffic engineered to take into consideration the consequences of a failure (i.e. by routing the backup LSP such that it can carry all of the traffic from the primary without affecting other LSPs), the load oscillations are no longer undesirable -- in fact, you don't care if the load shifts around in the network because you engineered for that eventuality.

So, what if we can use the backup path as the main path, though
previous path is restored.
Now, the load oscillations will be less.
What is your idea on this.
Is this worth enough? or is this ridiculous?

It really depends on how much of a problem it is to use the backup LSP on a given network. In a network with a lot of spare capacity, you probably don't care how optimal the LSPs are, so using the backup path as the new "primary" is OK. In a network where optimal LSP layout is important (and assuming you have a TE tool that can calculate optimal LSP paths) the backup path will always be sub-optimal because it's not the primary. So, given a longer term topological change to the network (i.e. a node or link out of service), the best solution is to re-run the optimization to figure out the new optimal LSP layout and re-route the LSPs accordingly.

So, no, it's not ridiculous.

Cheers,

Mathew

| Mathew Lodge                 | mathew@cplane.com     |
| Director, Product Management | Ph: +1 408 789 4068   |
| CPLANE, Inc.                 | http://www.cplane.com |