MPLS-TE tunnel not coming up

[Chanuka D]

Hi all,

We are having this topology in our network and I want to build a TE tunnel between two edge routers. For edge routers I followed the exact steps which was described in documentation. But TE tunnels are coming up. MPLS is functioning properly and do label switching. Loopbacks are reachable. Is there any additional configurations to be done in cisco router as well in this scenario ? What could cause TE tunnels not to come up from cisco router's perspective ?  Thanks. 

[smailmilak]

Hi,

provide us the output from this command
IOS: show mpls traffic-eng tunnels tunnel XX detail
XR: show mpls traffic-eng tunnels XX detail

[Chanuka D]

Hi  smailmilak,

crw-gw-1-sbjmen-pip.my#sh mpls traffic-eng tunnels

P2P TUNNELS/LSPs:

P2MP TUNNELS:

P2MP SUB-LSPS:
crw-gw-1-sbjmen-pip.my#

Here is the output requested. No info to display. 

[smailmilak]

Yeah, I see now that you are using Mikrotik routers. The output on the core device is empty because no MPLS TE tunnel is UP.

What you need for MPLS TE is to have OSPF/ISIS configured for MPLS TE, RSVP and labels end-to-end. Harold has posted all you need for proper MPLS TE config.

[Chanuka D]

Hi smailmilak,

Thank you very much for your response. I appreciate if you could explain what is RSVP how it works and how it integrates with MPLS TE. Please see my reply to Harold also. Thank you very much !

[smailmilak]

Best way to explain this is to copy/paste from the MPLS Fundamentals book written by Luc De Ghein.

RSVP uses PATH and RESV messages to signal a path. The TE head end router sends the PATH
messages to the tail end router, whereas the RESV messages take the exact but opposite path back
to the head end router. The head end router of a TE tunnel computes the best path that the TE
tunnel should take from the TE database, considering bandwidth and other constraints.

[Harold Ritter · Accepted solution]

Hi Chanuka,

You do indeed need some configuration in the Cisco equipment that is being used as a MPLS TE middle point router. For more information, please refer to the following documents:

Configuring MPLS Basic Traffic Engineering Using IS-IS:

https://www.cisco.com/c/en/us/support/docs/multiprotocol-label-switching-mpls/mpls/13737-mplsteisis.html

MPLS Basic Traffic Engineering Using OSPF Configuration Example

https://www.cisco.com/c/en/us/support/docs/multiprotocol-label-switching-mpls/mpls/13731-mpls-te-ospf.html

Regards,

[Chanuka D]

Hi Harold,

Thank you very much. You helped me again to overcome my problem. I already used following commands before start this discussion. 

(config)#mpls traffic-eng tunnels

(config-if)#mpls traffic-eng tunnels

(config-router)#mpls traffic-eng area 9

 (config-router)mpls traffic-eng router-id Loopback0

But the tunnel didn't come up. Then after referring the documentation you provided I found that I missed one command. 

(config-if)#ip rsvp bandwidth XXX XXX

Right after I entered this command under interface config my tunnel just came up. But I still have a one concern though. I reserved 300Mbps by using ip rsvp bandwidth command under my interfaces in cisco core router which is connecting to my edges. Can you please explain me how this command works ? This bandwidth is reserved for particular tunnel I created or all the tunnels that I'm gonna create in future as well ? Kindly explain. Thanks again. 

 

[Harold Ritter]

Hi Chanuka,

I am glad to hear the tunnel came up.

The "ip rsvp bandwitdh" command enables RSVP on the interface an specifies the bandwidth that can be reserved off that specific interface. If you just configured "ip rsvp bandwidth" without any value, 75% would be reserved by default.

Regards,

[Chanuka D]

Hi Harold,

Thank you for your explanation. Is it the total bandwidth I can use for any number of TE tunnels ? Let's say I reserve 750Mbps from a 1G interface by using ip rsvp bandwidth command. So I can use 750Mbps for TE tunnels and remaining 250 for normal forwarding. Is my understanding correct ? Kindly describe more if I'm wrong. Thanks.

[Harold Ritter]

Hi Chanuka,

Yes, the basic idea is that you would reserve 750M for all the TE tunnels reservation going through that specific interface and 250M for the non MPLS TE traffic. Bear in mind though that MPLS TE only makes BW reservation at the control plane level to ensure that you don't setup more TE Tunnels with specific BW reservation than available, but any tunnel can use more BW at the data plane level than what it reserved at the control plane level. It is up to the router at the tunnel head end to use rate limiting to ensure it doesn't send more traffic than what has been reserved.

Regards,

[Chanuka D]

Hi Harold,

Thank you very much for your clear explanation. I have one last question to ask from you. You mentioned that remaining bandwidth will be use for non MPLS TE traffic. Does that mean remaining bandwidth will be use for both IP traffic and normal MPLS forwading ? or you mean IP traffic only ? Thanks .

[Harold Ritter]

Hi Chanuka,

This does not specify what the remaining BW will be used for, but rather what portion of the interface BW should be used for MPLS TE tunnels BW reservation. So yes, technically you could use the remaning BW for the rest of your traffic, which could be MPLS, IP or anything else for that matter.

Regards,

[Chanuka D]

Hi Harold,

Thank you very much for sharing your knowledge with us. Appreciated !