A survey of network troubleshooting in MPLS

When the MPLS network is troubleshooting, it should follow from the control plane and the network plane to troubleshoot.

Let's take a look at what two planes do.

Above is the control plane, below is the network plane.

From the forwarding plane is the network plane, there will only be two kinds of messages, inbound for the IP packet, and inbound for the MPLS tag packet, then the corresponding two message forwarding will also be different, IP packet routing is posted in the forwarding, MPLS Group is Lfib--label forwarding Information base is being forwarded. Route forwarding is the routing table of the control plane. The only path to be elected according to the optimality principle is that the label forwarding is distributed according to the LDP agreement on the basis of the IGP.

Therefore, a complete MPLS forwarding, first of all, to ensure that the control plane of the IGP routing is normal work, and then to ensure that the LDP normal distribution tag, establish FEC, the last LFIB table correctly set up, in order to talk about the issue of forwarding.

First, the depth of the label.

To detect the failure of a tagged message at the data level, you first need to know how many labels should be on each link in the network that contains the tagged data message.

Here are some simple scenarios as examples:

0 ordinary IPv4 over mpls,1 a label. (4 bytes, LDP distribution)

0MPLS VPN or atom, or 6PE or 6vpe,2 tags (8 bytes, one is the inner layer of the label MP-BGP distribution, one is the outer label LDP allocation)

0 Flow Engineering 2 labels. (8 bytes, one is the label of the inner MP-BGP allocation, the other is the label of the outer RSVP assignment).

0 Fast Reroute FRR traffic engineering, 2 tags.

Now let's look at the TTL, the role that this important thing plays in the MPLS network. Let's review the normal IP TTL mode of operation by the way.

The following is the operating mode of an ordinary IP ttl: