[afnog] MPLS LSP UP but no traffic
ibtisam jamal
ibty.jamal at gmail.com
Mon Oct 17 09:40:17 UTC 2011
THanks Mark,
Interface configs and protocols is as below ,the whole network is in area 0
ospf
interfaces ge-0/1/1
vlan-tagging;
mtu 1600;
gigether-options {
no-auto-negotiation;
}
unit 0 {
vlan-id 98;
family inet {
address 10.87.8.198/30;
}
family mpls;
}
protocols bgp
hold-time 30;
graceful-restart;
group Core_BB_MPBGP {
type internal;
local-address 10.87.0.14;
family inet {
unicast;
}
family inet-vpn {
unicast;
}
peer-as 64514;
neighbor 10.87.0.1;
neighbor 10.87.0.2;
neighbor 10.87.0.3;
neighbor 10.87.0.4;
neighbor 10.87.0.5;
neighbor 10.87.0.6;
neighbor 10.87.0.19;
neighbor 10.87.0.20;
neighbor 10.87.0.11;
neighbor 10.87.0.12;
neighbor 10.87.0.13;
neighbor 10.87.0.15;
neighbor 10.87.0.16;
neighbor 10.87.0.17;
neighbor 10.87.0.18;
neighbor 10.87.0.21;
neighbor 10.87.0.22;
neighbor 10.87.0.23;
neighbor 10.87.0.24;
}
configuration protocols ospf
traceoptions {
file ospf.log size 1m files 20;
flag event detail;
flag hello detail;
flag error detail;
}
traffic-engineering;
reference-bandwidth 100g;
area 0.0.0.0 {
interface lo0.0 {
passive;
}
interface ge-0/1/3.206 {
inactive: bfd-liveness-detection {
minimum-interval 400;
multiplier 3;
}
}
interface ge-0/1/0.0 {
inactive: bfd-liveness-detection {
minimum-interval 400;
multiplier 3;
}
}
interface ge-0/1/1.0 {
bfd-liveness-detection {
minimum-interval 400;
multiplier 3;
}
}
}
Question: Is it OK or right to have LSPs that contain a mix of microwave and
fibre eg A-B fibre B-C is Microwave .Then you have an LSP from A-B-C ?
On Mon, Oct 17, 2011 at 11:49 AM, Mark Tinka <mtinka at globaltransit.net>wrote:
> On Monday, October 17, 2011 03:46:54 PM ibtisam jamal wrote:
>
> > > We have several links, one as the primary
> > > and others as
> > >
> > > secondary.The primary link is a fibre link and
> > > secondary links are microwave.Something odd has been
> > > happening on the primary link.It is physicaly up and
> > > lsp is up but it has no traffic and the secondary
> > > links do not take over because,rsvp knows that the lsp
> > > is up.I am trying to find out why we have this and how
> > > we can stop it from occurring again
>
> The configuration you've provided isn't sufficient on its
> own. Could you perhaps drop configurations on your
> [interfaces], and [protocols] section?
>
> It takes a mountain and all your dogs to get MPLS-TE right
> on a router (which is not to mean it's hard, but that a lot
> of components need to come together in harmony), so we'll
> need more information than you've given for now.
>
> > > also what is the impact of setting ERO options such as
> > > strict and loose ?
>
> Well, expanded loose hops are typically used in Inter-Area
> TE situations. By Inter-Area, we mean that you're running
> IS-IS with multiple levels or OSPF with multiple areas.
>
> In Inter-Area TE scenarios, the ingress routers have no
> knowledge of the "strict" hops beyond their area (the TED -
> TE Database - is retained only within a particular level or
> area by IS-IS or OSPF). This why the paths defined for the
> LSP are "loosely" specified at the ingress router.
>
> Otherwise, you would use a strict ERO provided your routers
> are all in the same IS-IS level or OSPF area.
>
> PS: note that unlike on Cisco routers, you cannot disable
> CSPF in Junos if you want to support expanded loose
> hops.
>
> Cheers,
>
> Mark.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://afnog.org/pipermail/afnog/attachments/20111017/7deb2188/attachment-0001.html>
More information about the afnog
mailing list