ccna ccie

ccna ccie

It periodically changes tunnel bandwidth(BW) reservation based on traffic out tunnel. Ip address 10.1.23.2 255.255.255.0 AutoRoute: disabled LockDown: disabled Loadshare: 0 bw-based auto-bw: disabled Level2 area Net 49.0001.1111.2222.3333.00 Ip unnumbered loopback 0 tunnel destination 5.5.5.5 tunnel mode mpls traffic-eng An IBGP adjacency is established between R1 and R4 . The IBGP adjacency is established on the physical interface. Established between R1 and R5 , R4 and R6 Tag rewrite with Fa0/0, 10.1.12.2, tags imposed: {203} Mpls traffic-eng tunnels mpls label range 300 399 *Aug 18 09:06:07.919: min unit=0 bytes, max pkt size=2147483647 bytes Maximum bandwidth : 12500000 !! Set MPLS TE routerID The default LDP label space is platform-based per-platform or device-based. What do you mean? Let's look at the B router, which bundles the label 34 for the prefix X , and publishes the label binding information to all LDP neighbors, which is the same for everyone, and everyone has a label 34 . This is based on the platform. Then there is an interface-based tag space. OSI network layer addressing is achieved by using two types of hierarchical addresses: NSAP and NET . Outgoing interface Tag or VC *Aug 18 09:06:07.919: Setup Prio: 7, Holding Prio: 7 Length of Value Interface loopback0 Router-id 4.4.4.4 Protocol overview LSPDB Reservable Bandwidth[7]: 0x2995 5.5.5.5/32 Link ID:: 0 Interface fast1/0 Router isis (Link ID) Designated Router address: 10.1.12.1 (Link Data) Router Interface address: 10.1.12.2 Number of MTID metrics: 0 MPLS TE tunnel IS-IS Redistribute ospf 1 vrf VPN-A match internal external 1 external 2 Network 10.1.12.1 0.0.0.0 area 0 The overload feature is, after all, a product of an era in which device performance is not strong enough. However , the extended application of overload is very worthy of attention. Let's take a look at the following picture: 17 AutoRoute: disabled LockDown: disabled Loadshare: 20000 bw-based auto-bw: disabled The unit of the OSI protocol stack defining data is the protocol data unit PDU , so the OSI view data frame is the PDU of the data link layer . The TE LSP tunnel is initiated by the head end of the tunnel. ! R1.00-00 *Aug 18 09:06:07.919: Network 10.1.23.2 0.0.0.0 area 0 SESSION The LSP overload-bit generated by this router is set. This will cause the router to not be used by other routers as the next hop to any destination, except for its locally connected network segment. This is generally used when the router is a leaf node,ccna ccie, that is, the end of the network, directly connected to the user. !! Activate IS-IS on the interface So far, the basic configuration has been completed. Now let 's create a TE Tunnel on R2 : LSP Checksum Regular area (non-backbone area) Options: (No TOS-capability, DC) LS Type: Opaque Area Link Overview Next-address 10.1.34.4 With PHP's penultimate hop pop-up mechanism, C allocates POP labels for local direct-connected prefixes and advertises them to other LDP neighbors. Configuration example 2 : Although R5 and R6 can learn each other's routes, they cannot communicate with each other because routing black holes appear on R2 and R3 . The solution is to use MPLS , we turn the Core into an MPLS domain: 2.2.2.2/32 The routing protocol in the IP network forwards the data packet according to the minimum metric principle. Next Hop Static route The tunnel 's build priority cannot be set higher than the hold priority. Look at the example above: A is in the area 49.0001 , B is in 49.0002 , because they are in different areas, so they become Level 2 adjacencies. Maintain adjacencies in the region at the same time. If A is configured: R4.00-00 For example, in the above topology, R1 establish a to R4 of TE the Tunnel , all routers have declared their loopback into OSPF . All interfaces cost are 1 . Path: valid 10.1.23.0/24 Router(config-if)#mpls traffic-eng flooding thresholds up/down ? Ip address 10.1.56.5 255.255.255.0 Hop Addr: 10.1.12.2 LIH: 0x0D000408 For example, the above figure, we can test on R2 , the configuration is as follows: Interface eth 0/1 A pseudonode represents a broadcast multi-access network (including all ISs connected to this LAN ), which actually "as a router" (above, gray router) in the network. In other places, this is a whole. So does this whole have to generate an LSP ? This LSP is generated by the DIS . 1/0/0 *Aug 18 09:06:07.919: version:1 flags:0000 cksum:7A42 ttl:255 reserved:0 length:132 Pop tag 2.2.2.2 Ip address 10.1.34.3 255.255.255.0 Record Route: Router ospf 1 Next Hop Further analysis of the metric problem. We focus on 55.55.55.55 , after the basic configuration above, currently in the routing table of R2 Label subobject: Flags 0x1, C-Type 1, Label 0 After completing the above configuration, we can check that the routing network is now interoperable. Labels: 200 RECORD_ROUTE RFC 3567 ! Start LSP ID : Start LSP ID . Indicates the first LSP ID of the LSP range described in the TLV field . Path removal IS-Extended R2.01 Redistribute isis ip level-2 into level-1 route-map test LSPID The configuration of R3 is as follows: Reservable BW (kbps) 202 The two important thresholds in the LSP remaining time field are: LSP maxage and LSP refresh interval Information, as shown below: Mpls traffic-eng tunnels ip rsvp bandwidth R1#show isis neighbors Not supported yet Router-id 3.3.3.3 Activate the fast-reroute feature on R1 and observe the phenomenon Interface loopback0 Interface Loopback0 Level2 , using the interface level command: isis circuit-type level2 , then R2 and R3 will not advertise these direct links when advertising LSPs in area 49.0001 . So the routing table for R1 becomes: 0/0/0 Let's take a look, why is it exposed to network information? Suppose A below router E , E made to be a packet to D under F. , The data packet is A tag the transferred B , assumed that the BC loop appears between subsequent label TTL is decremented to 0 , since the TTL value The setting is very small, so very quickly, the TTL is decremented to 0. At this time, B (message discarder) will send an ICMP error message back to E. Then, if B has E route (or label), B will The ICMP error message will be sent directly back to E. If not, B will continue to tag the error message and gradually pass it to D , D.As the PE , about E route, then back to the label and then loaded on E . Then, no matter what the situation, the final E can receive the ICMP error message generated by B.