IPWorld: CSPF

Constraint Shortest Path First (CSPF)

Without IGP traffic engineering extension, RSVP can only reserve bandwidth hop by hop. If at some point one of the nodes does not have enough available bandwidth, the LSP setup process just fails. Even if potentially there could be other alternative paths with enough bandwidth, RSVP on its own cannot be aware of them.

IGP Traffic Engineering extensions provides the solution to this specific problem, TE make RSVP aware of the full MPLS topology, including available bandwidth on each link, so that each LSR could make better use of network capacity.

The OSPF / IS-IS provide the traffic engineering database used by CSPF to calculate the best path according to the defined path constraints. The traffic engineering database provide routers all the information regarding available link bandwidths, link failures, or link colors (affinity) etc. With IGP TE routers distribute the MPLS topology and the information regarding available bandwidth. With TE database the endpoint can verify the path constraints e.g. bandwidth and establish the path according to the bandwidth availability even if it is not the best IGP path.

The following diagram shows the RSVP default behavior with applying traffic engineering extensions. Here RSVP followed the IGP best path to establish the LSP between ingress PE1 and egress PE2 routers.

Here the LSP is defined with bandwidth constraint

root@PE-1# show protocols mpls

no-cspf;

label-switched-path pe1-to-pe2 {

from 192.168.1.1;

to 192.168.1.2;

bandwidth 500m;

}

interface ge-0/0/8.0;

interface ge-0/0/9.0;

Following snippet showing the LSP path with require bandwidth 500 Mbps

Lets consider that LSP between PE1 and PE2 that is configured with bandwidth requirement of 500 mbps whereas only 100 mbps bandwidth available between LSR P1 and LSR P2 (in lab we can set the interface speed on P1 and P2 routers to 100 mbps). Without CSPF traffic engineering the RSVP try to establish LSP using the IGP best path but RSVP fail to established the LSP because the IGP best path use the link between routers P1 and P2 and the available bandwidth on that link is only 100 Mbps.

Although other paths (e.g. P3-P4) are available but RSVP can’t switch to other paths automatically because without traffic engineering RSVP can only follow the IGP best path as long as that path available and it remains the best path.

The above diagram shows the path failure as sufficient bandwidth is not available between the transit routers P1 and P2.

The following snippet shows the LSP status as down. It also shows the down reason as Required bandwidth unavailable.

The solution of this problem is to enable IGP traffic engineering and allow RSVP to use the traffic engineering database for path calculation. With this TE database the ingress LSR has the complete knowledge of available bandwidth at each hop therefor it can establish the path according to the available bandwidth.

In IS-IS the traffic engineering is enabled by default however with OSPF we need to enable it manually. Also, to allow MPLS RAVP to use CSPF we need to make sure that no-cspf command is not configured in mpls stanza.

Note: In order to have complete knowledge of the topology database OSPF traffic engineering must be enabled on all PE and P routers.

Configuration

root@PE-1# show protocols mpls

label-switched-path pe1-to-pe2 {

from 192.168.1.1;

to 192.168.1.2;

bandwidth 500m;

}

interface ge-0/0/8.0;

interface ge-0/0/9.0;

root@PE-1# show protocols ospf

traffic-engineering;

area 0.0.0.0 {

interface ge-0/0/9.0;

interface ge-0/0/8.0;

interface lo0.0 {

passive;

}

Following diagram shows the RSVP establish the path that full fill the bandwidth requirement. This time it establish the LSP with path PE1-P3-P4-PE2.