The MPLS-OPS Archive

[Ling Li <lli@axiowave.com>]

Hi All,

I just have a clarification question on "directly connected" and "adjacent"
to the previous node
mentioned in the email. Please see the following sample network:

RouterID:       10.0.10.1             10.0.10.2                   10.0.10.3
10.0.10.4
                  R1
--------------------R2--------------------------R3-------------------R4
Interface addr :   1   10.10.1.x/24  .2     .2   10.20.1.x/24  .1    .1
10.30.1.x/24  .2

Suppose R1 is ingress and R3 is egress of a LSP.

If  ERO sent from R1 is, 10.1.2/32 strict and 10.20.1.1 strict, everything
should be fine.

My question is if the following ERO sent from R1 are also valid?

1. 10.0.10.2/32 strict and 10.0.10.3/32 strict (RouterID in ERO)
2. 10.20.1.2/32 strict and 10.30.1.1/32 strict (not directly connected, but
is it
adjacent to the previous node?)

Have anyone tested these EROs on Juniper or Cisco?

Thanks,

Ling


-----Original Message-----
From: Joseph Anthony [mailto:tonyjoe20002002@yahoo.co.in]
Sent: Friday, January 31, 2003 3:40 AM
To: Sonali.Francis@lntinfotech.com; Wulf Losee
Cc: Alexandre Carlos; mpls-ops@mplsrc.com
Subject: Re: [MPLS-OPS]: Off-Topic doubt


Hello,
In simple terms, A strict hop must be directly connected (that is, adjacent)
to the previous node in the path. A loose hop is not necessarily directly
connected to the previous node; whether it is directly connected is unknown.

One area of usage, which I have come across is while configuring TE across
multiple OSPF areas on a Cisco implementation. You would need to configure
an explicit path and the path-option would need to specify the list of ABRs
the tunnel LSP must traverse as "loose" hops. 
Regards, 
Joe.

 Catch all the cricket action. Download Yahoo! Score tracker

-------
The MPLS-OPS Mailing List
Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
Archive: http://www.mplsrc.com/mpls-ops_archive.shtml