Hi all,
It's been a long time since I've done anything with MPLS and tunnels in
general, so I need the braintrust that is juniper-nsp to verify my thoughts on
something. I have a simple three router network as such:
R1--R2--R3
MPLS is configured on all core interfaces and an RSVP tunnel exists between R2
and R3 (and vice versa). Customer routes are configured to export to BGP, so
almost all routes on both R1 and R3 are seen as best path through the tunnel.
Ethernet MTU on the core interfaces is set to 9192, with MPLS MTU set to 1548,
and IP MTU set to 1500. I can't send anything through the tunnel larger than
1468-bytes.
Seeing's how my MTU settings on all core interfaces should leave plenty of
unnecessary headroom, would I be correct in my presumption that the underlying
circuit buildout from the carrier may not allow a physical media MTU over 1500?
There's a few reasons I came to this conclusion, but would like to know if I'm
missing anything blatantly obvious in my configs:
R1:
interfaces {
ge-0/0/0 {
mtu 9192;
unit 0 {
family inet {
mtu 1500;
address x.x.224.2/30;
}
family mpls {
mtu 1548;
}
}
}
}
protocols {
rsvp {
interface ge-0/0/0.0;
}
mpls {
path-mtu {
rsvp mtu-signaling;
}
label-switched-path to_r3 {
to x.x.207.3;
no-decrement-ttl;
}
interface ge-0/0/0.0;
}
}
R2:
interfaces {
ge-0/3/2 {
mtu 9192;
unit 0 {
family inet {
mtu 1500;
address x.x.224.1/30;
}
family mpls {
mtu 1548;
}
}
}
ge-0/3/3 {
vlan-tagging;
mtu 9192;
unit 0 {
vlan-id 1;
family inet {
mtu 1500;
address x.x.224.5/30;
}
family mpls {
mtu 1548;
}
}
}
}
protocols {
rsvp {
interface ge-0/3/2.0;
interface ge-0/3/3.0;
}
mpls {
interface ge-0/3/3.0;
interface ge-0/3/2.0;
}
}
R3:
interfaces {
ge-0/0/0 {
vlan-tagging;
mtu 9192;
unit 0 {
vlan-id 1;
family inet {
mtu 1500;
address x.x.224.6/30;
}
family mpls {
mtu 1548;
}
}
}
}
protocols {
rsvp {
interface ge-0/0/0.0;
}
mpls {
path-mtu {
rsvp mtu-signaling;
}
label-switched-path to_r2 {
to x.x.207.2;
no-decrement-ttl;
}
interface ge-0/0/0.0;
}
}
Thanks,
evt
_______________________________________________
juniper-nsp mailing list [email protected]
https://puck.nether.net/mailman/listinfo/juniper-nsp
