1、show ldp traffic-statistics(no config)| show log r7/ldp-statsshow ldp interfaceshow ldp neighborshow ldp sessionshow route table inet.3 -egress routers loopback addressshow route table mpls.0 -left:in label, right:swap label/popshow ldp database (Label-FEC)Notes: Automatically establish LSPs between
2、 the loopback addresses of all LDP-enabled routers(can be modified with the use of the transport-address keyword in the ldp stanza). By default, the JUNOS software implementation of LDP advertises a FEC for /32 interface routes only. LDP is reliance on the IGP to prevent loops means there is no need
3、 to implement some form of split horizon. LDP signaled LSPs are not displayed in the output of the show mpls lsp command (only RSVP listed) only the traffic associated with BGP next hops that resolve through inet.3 will be transported over an LSP (show route resolution)edit logical-routers r7 protoc
4、ols ldp - default metric is 1 in inet.3rootJNCIE# set track-igp-metric2. RSVP Signaled LSPsedit interfaces labr1# set fe-0/0/0.0 family mplsedit protocols mpls labr1# set interface alledit protocols rsvp - Enable RSVP Signalinglabr3# set interface all - recommend disabling fxp0show rsvp interfacesho
5、w rsvp neighbor -RSVP neighbors are detected onlywhen LSPs are actually signalededit protocols rsvplabr1# set interface fe-0/0/1 authentication-key jnxedit protocols mplslabr1# set label-switched-path r1-r7 to 10.0.9.7 bandwidth 10medit protocols mpls -with multiple TED domainslabr1# set label-switc
6、hed-path r1-r7 no-cspfshow rsvp sessionshow mpls lsp show route table mpls.0 label-switched-pathTroubleshoot:traceoptions file rsvp;flag error detail;flag path detail;flag pathtear detail;edit logical-routers r7 protocols mpls label-switched-path r7-r3 rootJNCIE# set metric 30 - default is IGPs best
7、 route metric in inet.33. ERO Constraintslabr2# set label-switched-path r2-r6 primary visit-r7-r5edit protocols mpls - EROs are considered strict by defaultlabr2# set path visit-r7-r5 10.0.4.9 10.0.9.7 loose loopback addresses should not be used in conjunction with strict hops, because by definition
8、 a strict hop should be directly connected. you can use a strict hop ERO pointing to a neighbors loopback interface when CSPF is used to compute the LSPs path.4. Link Coloringedit protocols mpls - color-to-numeric mappingslabr3# set admin-groups blue 4 (bit position).labr3# set interface so-0/2/0.10
9、0 admin-group blueedit protocols mpls label-switched-path r4-r3labr4# set admin-group include-any red exclude blueshow mpls lsp extensive (CSPF information)show ted database extensive | match nodeid|to|colorfile cspf;flag cspf;flag cspf-link;flag cspf-node; about CSPF Cant use CSPF(could use EROs) w
10、hen the netword have multiple routing domains(protocols) or Multi-Level IS-IS or Multi-Area OSPF,because TE information isnt leaked between IS-IS levels or OSPF areas JUNOS does offer support for LSP stitching CSPF include the computation of Fast Reroute detours, link bypass, LSP re-optimization, an
11、d the use of link coloring5. Installing Prefixes (precise control)edit protocols mpls label-switched-path r6-r1 -put in inet.3labr6# set install 10.0.5.254edit protocols mpls label-switched-path r6-r1 -move into inet.0labr6# set install 10.0.5/24 active with multiple equal-cost LSPs exist to the sam
12、e destination,default randomly balance traffic on a per-prefix basis by defaut,only LSP egress /32 prefix is installed in the inet.36. TE Shortcutslabel-switched-path r6-r3 to 10.0.3.3 -must in a single routing-domainedit protocols isis -automatic installation of all prefixes thatare considered down
13、stream of the LSP egress put into the inet.3 or inet.0 labr6# set traffic-engineering shortcutsedit protocols mpls -move all into inet.0 labr6# set traffic-engineering bgp-igp various other destinations also being mapped to the r6r3 LSP, this can be problematic(making the use of install generally pr
14、eferred). the LDP signaled LSP has also been moved into the inet.0edit protocols mpls -keep LDP/RSVP /32 prefix in inet.3 labr6# set traffic-engineering bgp-igp-both-ribsedit protocols mpls - = Routing Use Only, # = Forwarding Use Onlykeep IGP routes being active for routing policy use labr6# set tr
15、affic-engineering mpls-forwarding 7. Prefix-to-LSP Mappinga) forwarding table policy at the LSP ingress node edit logical-routers r7 policy-options policy-statement lsp-mapterm 1 from protocol bgp; neighbor 10.0.3.3; route-filter 0.0.0.0/0 upto /24; then install-nexthop lsp r7-r3; accept;term 2 inst
16、all-nexthop lsp r7-r3-prime; labr7# set forwarding-table export lsp-mapb) BGP next hop manipulation at the LSP egress node listing r3s lo0 interface under its area 0 definition, which results in the advertisement of both the primary and secondary lo0 addresses. a secondary lo0 address will not be ad
17、vertised into the TED,so disable cspf IBGP export policy will be used at the LSP egress node to selectively set the BGP next hop to the desired LSP egress address8. Secondary Pathsedit protocols mpls label-switched-path r1-r7 -establish the secondary path before the primary path failedlabr1# set sec
18、ondary r1-r7-prime standbyedit protocols mpls label-switched-path r1-r7 - change the RSVP reservation style from the default Fixed Filter (FF) to a Shared Explicit (SE) stylelabr1# set adaptive can define many secondary paths,but only one established a time define only secondary paths to prevent rev
19、erting back to primary(not switch among secondary paths)9. Fast Reroute and Link Protectionedit protocols mpls label-switched-path r6-r4 -protect the entire LSP,use TED to locate detour paths while regardless of CSPFlabr6# set fast-reroute-edit protocols mpls label-switched-path r6-r4 -protect a spe
20、cific interface,makes use of a TED and the CSPFlabr6# set link-protectionedit protocols rsvp labr3# set interface so-0/2/0.100 link-protection (path -ERO)show rsvp session ingress detail(on failed router)A key point to consider when dealing with both Fast Reroute and link protection is the need for
21、a TED with sufficient information to allow a successful CSPF calculation10. Preemption (setup and hold priority)edit protocols mpls label-switched-path r5-r1 labr5# set priority x x -default 7(weak) 0(strong)10. Control-TTLedit protocols mpls label-switched-path r2-r6 -only at ingresslabr2# set no-d
22、ecrement-ttl edit protocols mpls -at all nodes along the LSP,support multi-vendorslabr2# set no-propagate-ttl 11. Validate the data plane without using external prefixes or modifying default routing table integration.edit interfaces - the MPLS ping mechanism always targets a destination address of 1
23、27.0.0.1labr1# set lo0.0 family inet address 127.0.0.1ping mpls rsvp r5-r112. Ultimate hop poppingedit protocols ldp/mpls -why? patibility 2.CoSlabr6# set explicit-null13. Automatica bandwidth reservationlabr4# set statistics file mpls-stats set statistics auto-bandwidthedit protocols mpls label-swi
24、tched-path r4-r3 -default 86400slabr4# set auto-bandwidth minimum-bandwidth 1m adjust-interval 30014. RSVP-based LDP tunnelingedit protocols mpls label-switched-path r1-r7 -labr1#set ldp-tunnelingedit protocols ldp -labr1#set interface lo0.0Layer 3 VPNs (2547bis)1. Static and BGP routing for PE-CE c
25、onnectionedit routing-instances c2 labr6# set instance-type vrfedit routing-instances c2 -connect to celabr6# set interface fe-0/1/3edit routing-instances c2 -default VRF policy:automaticlly advertise and receive ; the routes from CE as well as the vrfs static and direct routes; local routes cant be
26、 exported;labr6# set vrf-target target:65412:420edit routing-options -automaticlabr6# set route-distinguisher-id 10.0.9.6edit routing-instances c2 -manuallabr6# set route-distinguisher as:nn or ip:nnedit routing-instances c2 routing-optionslabr6#set static route 220.220/16 next-hop 172.16.0.10 ping
27、220.220.0.1 routing-instance c2edit routing-instances c2 protocol bgp group c1labr4#set type external peer-as 65010 neighbor 172.16.0.6show bgp summary instance c1show route table c1edit routing-instances c2 -perform end-to-end ping and traceroutelabr6#set vrf-table-label2. MP-BGP among PE-PEedit protocols bgp group int neighbor 10.0.9.6 -both address family enabledlabr4# show family inet unicast;family inet-vpn show route table bgp.l3vpn.0show rout
copyright@ 2008-2023 冰点文库 网站版权所有
经营许可证编号:鄂ICP备19020893号-2