diff options
| -rw-r--r-- | docs/configuration/protocols/mpls.md | 812 |
1 files changed, 650 insertions, 162 deletions
diff --git a/docs/configuration/protocols/mpls.md b/docs/configuration/protocols/mpls.md index 71b14be2..3f734138 100644 --- a/docs/configuration/protocols/mpls.md +++ b/docs/configuration/protocols/mpls.md @@ -1,285 +1,773 @@ +--- +myst: + html_meta: + description: | + Multiprotocol Label Switching (MPLS) is a packet-forwarding method + that forwards traffic using exact-match lookups on MPLS labels + rather than IP address lookups. + keywords: mpls, ldp, label distribution protocol, lsp, label switching +--- + (mpls)= # MPLS -{abbr}`MPLS (Multi-Protocol Label Switching)` is a packet forwarding paradigm -which differs from regular IP forwarding. Instead of IP addresses being used to -make the decision on finding the exit interface, a router will instead use an -exact match on a 32 bit/4 byte header called the MPLS label. This label is -inserted between the ethernet (layer 2) header and the IP (layer 3) header. -One can statically or dynamically assign label allocations, but we will focus -on dynamic allocation of labels using some sort of label distribution protocol -(such as the aptly named Label Distribution Protocol / LDP, Resource Reservation -Protocol / RSVP, or Segment Routing through OSPF/ISIS). These protocols allow -for the creation of a unidirectional/unicast path called a labeled switched -path (initialized as LSP) throughout the network that operates very much like -a tunnel through the network. An easy way of thinking about how an MPLS LSP -actually forwards traffic throughout a network is to think of a GRE tunnel. -They are not the same in how they operate, but they are the same in how they -handle the tunneled packet. It would be good to think of MPLS as a tunneling -technology that can be used to transport many different types of packets, to -aid in traffic engineering by allowing one to specify paths throughout the -network (using RSVP or SR), and to generally allow for easier intra/inter -network transport of data packets. - -For more information on how MPLS label switching works, please go visit -[Wikipedia (MPLS)]. - -:::{note} -MPLS support in VyOS is not finished yet, and therefore its -functionality is limited. Currently there is no support for MPLS enabled VPN -services such as L2VPNs and mVPNs. RSVP support is also not present as the -underlying routing stack (FRR) does not implement it. Currently VyOS -implements LDP as described in RFC 5036; other LDP standard are the -following ones: RFC 6720, RFC 6667, RFC 5919, RFC 5561, RFC 7552, RFC 4447. -Because MPLS is already available (FRR also supports RFC 3031). -::: - -## Label Distribution Protocol - -The {abbr}`MPLS (Multi-Protocol Label Switching)` architecture does not assume -a single protocol to create MPLS paths. VyOS supports the Label Distribution -Protocol (LDP) as implemented by FRR, based on {rfc}`5036`. - -{abbr}`LDP (Label Distribution Protocol)` is a TCP based MPLS signaling protocol -that distributes labels creating MPLS label switched paths in a dynamic manner. -LDP is not a routing protocol, as it relies on other routing protocols for -forwarding decisions. LDP cannot bootstrap itself, and therefore relies on said -routing protocols for communication with other routers that use LDP. - -In order to allow for LDP on the local router to exchange label advertisements -with other routers, a TCP session will be established between automatically -discovered and statically assigned routers. LDP will try to establish a TCP -session to the **transport address** of other routers. Therefore for LDP to -function properly please make sure the transport address is shown in the -routing table and reachable to traffic at all times. - -It is highly recommended to use the same address for both the LDP router-id and -the discovery transport address, but for VyOS MPLS LDP to work both parameters -must be explicitly set in the configuration. - -Another thing to keep in mind with LDP is that much like BGP, it is a protocol -that runs on top of TCP. It however does not have an ability to do something -like a refresh capability like BGPs route refresh capability. Therefore one -might have to reset the neighbor for a capability change or a configuration -change to work. - -## Configuration Options +{abbr}`MPLS (Multiprotocol Label Switching)` is a packet-forwarding method +that differs from traditional IP forwarding. Instead of performing IP address +lookups, MPLS routers forward traffic using exact-match lookups on MPLS labels +inserted between the Layer 2 (Ethernet) and Layer 3 (IP) headers. + +MPLS labels can be assigned statically or dynamically. This section focuses on +dynamic label allocation using label distribution protocols such as the +{abbr}`LDP (Label Distribution Protocol)` and +{abbr}`RSVP-TE (Resource Reservation Protocol for Traffic Engineering)`, or +via {abbr}`SR (Segment Routing)` extensions to OSPF or IS-IS. + +These protocols establish a unidirectional path called a +{abbr}`LSP (Label-Switched Path)` across the network. When a packet enters +the MPLS network, the ingress router pushes an MPLS label onto it based on +the destination and forwards the packet along the LSP. Each intermediate +router uses the incoming label to look up the outgoing interface and next +label value, swaps the label, and forwards the packet on that interface. The +egress router pops the label and forwards the original packet using normal +IP routing. + +## MPLS support in VyOS + +MPLS support in VyOS is under development, so its feature set is currently +limited. Basic MPLS forwarding is functional, as the underlying FRR routing +stack supports [RFC 3031](https://datatracker.ietf.org/doc/html/rfc3031). + +Supported features: + +- **LDP**: Implemented according to + [RFC 5036](https://datatracker.ietf.org/doc/html/rfc5036). + +```{note} +Other related LDP standards include RFCs +[6720](https://datatracker.ietf.org/doc/html/rfc6720), +[6667](https://datatracker.ietf.org/doc/html/rfc6667), +[5919](https://datatracker.ietf.org/doc/html/rfc5919), +[5561](https://datatracker.ietf.org/doc/html/rfc5561), and +[7552](https://datatracker.ietf.org/doc/html/rfc7552). +``` + +Current limitations: + +- **MPLS VPNs**: No support yet for MPLS-enabled VPN services, such as + L2VPNs and mVPNs. +- **RSVP-TE**: Not supported, as the underlying FRR stack does not currently + implement it. + +## Label Distribution Protocol (LDP) + +MPLS supports various protocols for path creation. VyOS implements LDP via +the underlying FRR routing stack, per +[RFC 5036](https://datatracker.ietf.org/doc/html/rfc5036). + +LDP is a TCP-based MPLS signaling protocol that dynamically distributes +labels to create label-switched paths. It is not a routing protocol and +depends on existing routing protocols for forwarding decisions and +communication with other LDP-enabled routers. + +To exchange label advertisements, LDP establishes TCP sessions with +automatically discovered or statically configured neighbors. Each session is +formed using the peer's transport address, which must be present in the routing +table and continuously reachable. + +```{note} +It is recommended to use the same address for both the LDP router +ID and the discovery transport address. For MPLS LDP to function in VyOS, +both parameters must be explicitly configured. +``` + +```{note} +LDP has no mechanism to refresh an existing session. To apply a +configuration change or a change to negotiated session capabilities, the +LDP neighbor must be reset on the local router. +``` + +## Configuration ```{cfgcmd} set protocols mpls interface \<interface\> -Use this command to enable MPLS processing on the interface you define. +**Enable MPLS on the specified interface.** +``` + +Example: + +```none +set protocols mpls interface eth1 +``` + +```{cfgcmd} set protocols mpls parameters no-propagate-ttl + +**Disable propagation of the IP TTL into the MPLS header at label +imposition.** ``` +Example: + +```none +set protocols mpls parameters no-propagate-ttl +``` + +```{cfgcmd} set protocols mpls parameters maximum-ttl \<1-255\> + +**Configure the maximum TTL value for MPLS packets.** + +This value is used when the TTL is not propagated from the IP header. +The default is 255. +``` + +Example: + +```none +set protocols mpls parameters maximum-ttl 15 +``` ```{cfgcmd} set protocols mpls ldp interface \<interface\> -Use this command to enable LDP on the interface you define. +**Enable LDP on the specified interface.** +``` + +Example: + +```none +set protocols mpls ldp interface eth1 +``` + +```{cfgcmd} set protocols mpls ldp interface \<interface\> disable-establish-hello + +**Disable the triggered Hello that the router sends on the specified +interface in response to receiving an LDP Hello from a peer.** ``` +Example: + +```none +set protocols mpls ldp interface eth0 disable-establish-hello +``` ```{cfgcmd} set protocols mpls ldp router-id \<address\> -Use this command to configure the IP address used as the LDP router-id of the -local device. +**Configure the IP address used as the LDP router ID for the local device.** ``` +Example: + +```none +set protocols mpls ldp router-id 192.0.2.1 +``` ```{cfgcmd} set protocols mpls ldp discovery transport-ipv4-address \<address\> +**Configure the IPv4 transport address for IPv4 LDP connections.** +``` + +Example: + +```none +set protocols mpls ldp discovery transport-ipv4-address 192.0.2.1 ``` + ```{cfgcmd} set protocols mpls ldp discovery transport-ipv6-address \<address\> -Use this command to set the IPv4 or IPv6 transport-address used by LDP. +**Configure the IPv6 transport address for IPv6 LDP connections.** +``` + +Example: + +```none +set protocols mpls ldp discovery transport-ipv6-address 2001:db8::1 +``` + +```{cfgcmd} set protocols mpls ldp neighbor \<ipv4-address\> password \<password\> + +**Configure authentication for the LDP session with the specified +neighbor.** +``` + +```{note} +For the session to establish successfully, both peers must be configured +with the same password. +``` + +Example: + +```none +set protocols mpls ldp neighbor 192.0.2.2 password mysharedsecret +``` + +```{cfgcmd} set protocols mpls ldp neighbor \<ipv4-address\> session-holdtime \<15-65535\> + +**Configure the LDP session hold time, in seconds, advertised to the +specified neighbor.** + +The neighbor must be reset for this change to take effect. +``` + +Example: + +```none +set protocols mpls ldp neighbor 192.0.2.2 session-holdtime 180 +``` + +```{cfgcmd} set protocols mpls ldp neighbor \<ipv4-address\> ttl-security \<disable | 1-254\> + +**Configure TTL security ({abbr}`GTSM (Generalized TTL Security Mechanism)`) +for the LDP session with the specified neighbor.** + +Use `disable` to turn off TTL security, or set a hop-count value (1-254) to +restrict which incoming packets are accepted based on their TTL. +``` + +Example: + +```none +set protocols mpls ldp neighbor 192.0.2.2 ttl-security 5 +``` + +```{cfgcmd} set protocols mpls ldp discovery hello-ipv4-interval \<1-65535\> + +**Configure the interval, in seconds, at which the router sends IPv4 LDP +Hello messages.** +``` + +Example: + +```none +set protocols mpls ldp discovery hello-ipv4-interval 5 +``` + +```{cfgcmd} set protocols mpls ldp discovery hello-ipv4-holdtime \<1-65535\> + +**Configure the hold time, in seconds, advertised in IPv4 LDP Hello +messages.** +``` + +Example: + +```none +set protocols mpls ldp discovery hello-ipv4-holdtime 15 +``` + +```{cfgcmd} set protocols mpls ldp discovery hello-ipv6-interval \<1-65535\> + +**Configure the interval, in seconds, at which the router sends IPv6 LDP +Hello messages.** +``` + +Example: + +```none +set protocols mpls ldp discovery hello-ipv6-interval 5 +``` + +```{cfgcmd} set protocols mpls ldp discovery hello-ipv6-holdtime \<1-65535\> + +**Configure the hold time, in seconds, advertised in IPv6 LDP Hello +messages.** +``` + +Example: + +```none +set protocols mpls ldp discovery hello-ipv6-holdtime 15 +``` + +```{cfgcmd} set protocols mpls ldp discovery session-ipv4-holdtime \<15-65535\> + +**Configure the LDP session hold time, in seconds, advertised in IPv4 LDP +Initialization messages.** +``` + +Example: + +```none +set protocols mpls ldp discovery session-ipv4-holdtime 180 +``` + +```{cfgcmd} set protocols mpls ldp discovery session-ipv6-holdtime \<15-65535\> + +**Configure the LDP session hold time, in seconds, advertised in IPv6 LDP +Initialization messages.** +``` + +Example: + +```none +set protocols mpls ldp discovery session-ipv6-holdtime 180 +``` + +```{cfgcmd} set protocols mpls ldp import ipv4 import-filter filter-access-list \<1-2699\> + +**Filter the IPv4 label bindings (associations between an IPv4 prefix and +an MPLS label) from LDP peers using the specified access list.** + +Only label bindings whose prefix is permitted by the access list are +accepted. +``` + +Example: + +```none +set protocols mpls ldp import ipv4 import-filter filter-access-list 10 +``` + +```{cfgcmd} set protocols mpls ldp import ipv4 import-filter neighbor-access-list \<1-2699\> + +**Filter the IPv4 label bindings (associations between an IPv4 prefix +and an MPLS label) from LDP peers using the specified access list.** + +Only label bindings received from a peer whose IPv4 address is permitted +by the access list are accepted. +``` + +Example: + +```none +set protocols mpls ldp import ipv4 import-filter neighbor-access-list 20 +``` + +```{cfgcmd} set protocols mpls ldp import ipv6 import-filter filter-access-list6 \<1-2699\> + +**Filter the IPv6 label bindings (associations between an IPv6 prefix and +an MPLS label) from LDP peers using the specified access list.** + +Only label bindings whose prefix is permitted by the access list are +accepted. +``` + +Example: + +```none +set protocols mpls ldp import ipv6 import-filter filter-access-list6 10 +``` + +```{cfgcmd} set protocols mpls ldp import ipv6 import-filter neighbor-access-list6 \<1-2699\> + +**Filter the IPv6 label bindings (associations between an IPv6 prefix +and an MPLS label) from LDP peers using the specified access list.** + +Only label bindings received from a peer whose IPv6 address is permitted +by the access list are accepted. +``` + +Example: + +```none +set protocols mpls ldp import ipv6 import-filter neighbor-access-list6 20 +``` + +```{cfgcmd} set protocols mpls ldp export ipv4 export-filter filter-access-list \<1-2699\> + +**Filter the IPv4 label bindings (associations between an IPv4 prefix and +an MPLS label) advertised to LDP peers using the specified access list.** + +Only label bindings whose prefix is permitted by the access list are +advertised. +``` + +Example: + +```none +set protocols mpls ldp export ipv4 export-filter filter-access-list 10 +``` + +```{cfgcmd} set protocols mpls ldp export ipv4 export-filter neighbor-access-list \<1-2699\> + +**Filter the IPv4 label bindings (associations between an IPv4 prefix +and an MPLS label) advertised to LDP peers using the specified access +list.** + +Label bindings are advertised only to peers whose IPv4 address is +permitted by the access list. +``` + +Example: + +```none +set protocols mpls ldp export ipv4 export-filter neighbor-access-list 20 +``` + +```{cfgcmd} set protocols mpls ldp export ipv6 export-filter filter-access-list6 \<1-2699\> + +**Filter the IPv6 label bindings (associations between an IPv6 prefix and +an MPLS label) advertised to LDP peers using the specified access list.** + +Only label bindings whose prefix is permitted by the access list are +advertised. ``` -```{cfgcmd} set protocols mpls ldp neighbor \<address\> password \<password\> +Example: -Use this command to configure authentication for LDP peers. Set the -IP address of the LDP peer and a password that should be shared in -order to become neighbors. +```none +set protocols mpls ldp export ipv6 export-filter filter-access-list6 10 +``` + +```{cfgcmd} set protocols mpls ldp export ipv6 export-filter neighbor-access-list6 \<1-2699\> + +**Filter the IPv6 label bindings (associations between an IPv6 prefix +and an MPLS label) advertised to LDP peers using the specified access +list.** + +Label bindings are advertised only to peers whose IPv6 address is +permitted by the access list. +``` + +Example: + +```none +set protocols mpls ldp export ipv6 export-filter neighbor-access-list6 20 ``` -```{cfgcmd} set protocols mpls ldp neighbor \<address\> session-holdtime \<seconds\> +```{cfgcmd} set protocols mpls ldp export ipv4 explicit-null -Use this command to configure a specific session hold time for LDP peers. -Set the IP address of the LDP peer and a session hold time that should be -configured for it. You may have to reset the neighbor for this to work. +**Configure the router to advertise the Explicit Null label for its +directly connected IPv4 prefixes to LDP neighbors.** + +This instructs the penultimate-hop router to forward labeled packets +unchanged without removing the MPLS label. ``` -```{cfgcmd} set protocols mpls ldp neighbor \<address\> ttl-security \<disable | hop count\> +Example: -Use this command to enable, disable, or specify hop count for TTL security -for LDP peers. By default the value is set to 255 (or max TTL). +```none +set protocols mpls ldp export ipv4 explicit-null ``` -```{eval-rst} -.. cfgcmd:: set protocols mpls ldp discovery hello-ipv4-interval <seconds> -.. cfgcmd:: set protocols mpls ldp discovery hello-ipv4-holdtime <seconds> -.. cfgcmd:: set protocols mpls ldp discovery hello-ipv6-interval <seconds> -.. cfgcmd:: set protocols mpls ldp discovery hello-ipv6-holdtime <seconds> +```{cfgcmd} set protocols mpls ldp export ipv6 explicit-null + +**Configure the router to advertise the Explicit Null label for its +directly connected IPv6 prefixes to LDP neighbors.** - Use these commands if you would like to set the discovery hello and hold time - parameters. +This instructs the penultimate-hop router to forward labeled packets +unchanged without removing the MPLS label. ``` -```{eval-rst} -.. cfgcmd:: set protocols mpls ldp discovery session-ipv4-holdtime <seconds> -.. cfgcmd:: set protocols mpls ldp discovery session-ipv6-holdtime <seconds> +Example: - Use this command if you would like to set the TCP session hold time intervals. +```none +set protocols mpls ldp export ipv6 explicit-null ``` -```{eval-rst} -.. cfgcmd:: set protocols mpls ldp import ipv4 import-filter filter-access-list - <access list number> -.. cfgcmd:: set protocols mpls ldp import ipv6 import-filter filter-access-list6 - <access list number> +```{cfgcmd} set protocols mpls ldp allocation ipv4 access-list \<1-2699\> + +**Restrict MPLS label allocation to IPv4 prefixes permitted by the +specified access list.** - Use these commands to control the importing of forwarding equivalence classes - (FECs) for LDP from neighbors. This would be useful for example on only - accepting the labeled routes that are needed and not ones that are not - needed, such as accepting loopback interfaces and rejecting all others. +By default, LDP allocates a label for every IPv4 prefix in the routing +table. A common best practice is to limit allocation to loopback +addresses. ``` -```{eval-rst} -.. cfgcmd:: set protocols mpls ldp export ipv4 export-filter filter-access-list - <access list number> -.. cfgcmd:: set protocols mpls ldp export ipv6 export-filter filter-access-list6 - <access list number> +Example: - Use these commands to control the exporting of forwarding equivalence classes - (FECs) for LDP to neighbors. This would be useful for example on only - announcing the labeled routes that are needed and not ones that are not - needed, such as announcing loopback interfaces and no others. +```none +set protocols mpls ldp allocation ipv4 access-list 10 ``` -```{eval-rst} -.. cfgcmd:: set protocols mpls ldp export ipv4 explicit-null -.. cfgcmd:: set protocols mpls ldp export ipv6 explicit-null +```{cfgcmd} set protocols mpls ldp allocation ipv6 access-list6 \<1-2699\> + +**Restrict MPLS label allocation to IPv6 prefixes permitted by the +specified access list.** - Use this command if you would like for the router to advertise FECs with a - label of 0 for explicit null operations. +By default, LDP allocates a label for every IPv6 prefix in the routing +table. A common best practice is to limit allocation to loopback +addresses. ``` -```{eval-rst} -.. cfgcmd:: set protocols mpls ldp allocation ipv4 access-list <access list number> -.. cfgcmd:: set protocols mpls ldp allocation ipv6 access-list6 <access list number> +Example: - Use this command if you would like to control the local FEC allocations for - LDP. A good example would be for your local router to not allocate a label for - everything. Just a label for what it's useful. A good example would be just a - loopback label. +```none +set protocols mpls ldp allocation ipv6 access-list6 10 ``` ```{cfgcmd} set protocols mpls ldp parameters cisco-interop-tlv -Use this command to use a Cisco non-compliant format to send and interpret -the Dual-Stack capability TLV for IPv6 LDP communications. This is related to -{rfc}`7552`. +**Configure the router to negotiate the Dual-Stack capability +{abbr}`TLV (Type-Length-Value)` +([RFC 7552](https://datatracker.ietf.org/doc/html/rfc7552)) using a Cisco +non-compliant format for dual-stack LDP sessions.** +``` + +Example: + +```none +set protocols mpls ldp parameters cisco-interop-tlv ``` ```{cfgcmd} set protocols mpls ldp parameters ordered-control -Use this command to use ordered label distribution control mode. FRR -by default uses independent label distribution control mode for label -distribution. This is related to {rfc}`5036`. +**Enable LDP Ordered Label Distribution Control mode +([RFC 5036](https://datatracker.ietf.org/doc/html/rfc5036)) for the +router.** + +By default, the router operates in Independent Label Distribution Control +mode. +``` + +Example: + +```none +set protocols mpls ldp parameters ordered-control ``` ```{cfgcmd} set protocols mpls ldp parameters transport-prefer-ipv4 -Use this command to prefer IPv4 for TCP peer transport connection for LDP -when both an IPv4 and IPv6 LDP address are configured on the same interface. +**Configure the router to prefer an IPv4 TCP transport connection for LDP +peering when LDP is configured for dual-stack operation (both IPv4 and +IPv6 address families are enabled).** +``` + +Example: + +```none +set protocols mpls ldp parameters transport-prefer-ipv4 ``` ```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv4 enable + +**Enable the router to accept targeted IPv4 LDP sessions.** + +This allows the router to establish LDP adjacencies with non-directly +connected peers by responding to inbound targeted (unicast) LDP Hello +messages. +``` + +Example: + +```none +set protocols mpls ldp targeted-neighbor ipv4 enable ``` ```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv6 enable -Use this command to enable targeted LDP sessions to the local router. The -router will then respond to any sessions that are trying to connect to it that -are not a link local type of TCP connection. +**Enable the router to accept targeted IPv6 LDP sessions.** + +This allows the router to establish LDP adjacencies with non-directly +connected peers by responding to inbound targeted (unicast) LDP Hello +messages. +``` + +Example: + +```none +set protocols mpls ldp targeted-neighbor ipv6 enable ``` ```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv4 address \<address\> + +**Configure the router to initiate a targeted IPv4 LDP session with the +specified remote {abbr}`LSR (Label Switching Router)`.** +``` + +Example: + +```none +set protocols mpls ldp targeted-neighbor ipv4 address 192.0.2.5 ``` ```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv6 address \<address\> -Use this command to enable the local router to try and connect with a targeted -LDP session to another router. +**Configure the router to initiate a targeted IPv6 LDP session with the +specified remote LSR.** +``` + +Example: + +```none +set protocols mpls ldp targeted-neighbor ipv6 address 2001:db8::5 +``` + +```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv4 hello-holdtime \<1-65535\> + +**Configure the hold time, in seconds, advertised in targeted IPv4 LDP +Hello messages.** +``` + +Example: + +```none +set protocols mpls ldp targeted-neighbor ipv4 hello-holdtime 45 ``` -```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv4 hello-holdtime \<seconds\> +```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv4 hello-interval \<1-65535\> + +**Configure the interval, in seconds, between targeted IPv4 LDP Hello +messages sent to remote LSRs.** ``` -```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv4 hello-interval \<seconds\> +Example: + +```none +set protocols mpls ldp targeted-neighbor ipv4 hello-interval 15 +``` + +```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv6 hello-holdtime \<1-65535\> + +**Configure the hold time, in seconds, advertised in targeted IPv6 LDP +Hello messages.** ``` -```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv6 hello-holdtime \<seconds\> +Example: + +```none +set protocols mpls ldp targeted-neighbor ipv6 hello-holdtime 45 ``` -```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv6 hello-interval \<seconds\> +```{cfgcmd} set protocols mpls ldp targeted-neighbor ipv6 hello-interval \<1-65535\> -Use these commands if you would like to set the discovery hello and hold time -parameters for the targeted LDP neighbors. +**Configure the interval, in seconds, between targeted IPv6 LDP Hello +messages sent to remote LSRs.** ``` -### Sample configuration to setup LDP on VyOS +Example: ```none -set protocols ospf area 0 network '192.168.255.252/32' <--- Routing for loopback -set protocols ospf area 0 network '192.168.0.5/32' <--- Routing for an interface connecting to the network -set protocols ospf parameters router-id '192.168.255.252' <--- Router ID setting for OSPF -set protocols mpls interface 'eth1' <--- Enable MPLS for an interface connecting to network -set protocols mpls ldp discovery transport-ipv4-address '192.168.255.252' <--- Transport address for LDP for TCP sessions to connect to -set protocols mpls ldp interface 'eth1' <--- Enable LDP for an interface connecting to network -set protocols mpls ldp interface 'lo' <--- Enable LDP on loopback for future services connectivity -set protocols mpls ldp router-id '192.168.255.252' <--- Router ID setting for LDP -set interfaces ethernet eth1 address '192.168.0.5/31' <--- Interface IP for connecting to network -set interfaces loopback lo address '192.168.255.252/32' <--- Interface loopback IP for router ID and other uses +set protocols mpls ldp targeted-neighbor ipv6 hello-interval 15 ``` -## Operational Mode Commands +## Operation -When LDP is working, you will be able to see label information in the outcome -of `show ip route`. Besides that information, there are also specific *show* -commands for LDP: +When LDP is established, you can view label assignments directly within the +`show ip route` or `show ipv6 route` outputs. To inspect LDP operational +states, use the following commands. ### Show ```{opcmd} show mpls ldp binding -Use this command to see the Label Information Base. +Show the {abbr}`LIB (Label Information Base)`. +``` + +```{opcmd} show mpls ldp binding \<prefix\> + +Show LDP label bindings for the specified prefix. +The `<prefix>` parameter accepts an IPv4 prefix (`x.x.x.x/x`) or IPv6 +prefix (`h:h:h:h:h:h:h:h/h`). +``` + +```{opcmd} show mpls ldp binding detail + +Show detailed LDP Label Information Base (LIB) entries. +``` + +```{opcmd} show mpls ldp binding neighbor \<address\> + +Show LDP label bindings learned from the specified peer. + +The `<address>` parameter accepts an IPv4 or IPv6 address. +``` + +```{opcmd} show mpls ldp binding local-label \<label\> + +Show LDP label bindings associated with the specified locally +assigned label value. +``` + +```{opcmd} show mpls ldp binding remote-label \<label\> + +Show LDP label bindings associated with the specified remotely +assigned label value. ``` ```{opcmd} show mpls ldp discovery -Use this command to see discovery hello information +Show LDP discovery Hello information. +``` + +```{opcmd} show mpls ldp discovery detail + +Show detailed LDP discovery Hello information. ``` ```{opcmd} show mpls ldp interface -Use this command to see LDP interface information +Show LDP interface information. ``` ```{opcmd} show mpls ldp neighbor -Use this command to see LDP neighbor information +Show basic LDP neighbor information. ``` ```{opcmd} show mpls ldp neighbor detail -Use this command to see detailed LDP neighbor information +Show detailed LDP neighbor information. +``` + +```{opcmd} show mpls ldp neighbor capabilities + +Show the LDP capabilities negotiated with each peer during session +initialization. +``` + +```{opcmd} show mpls ldp neighbor \<address\> + +Show LDP neighbor information for the specified peer. + +The `<address>` parameter accepts an IPv4 or IPv6 address. +``` + +```{opcmd} show mpls ldp neighbor \<address\> detail + +Show detailed LDP session information for the specified peer. +``` + +```{opcmd} show mpls ldp neighbor \<address\> capabilities + +Show the LDP capabilities negotiated with the specified peer. +``` + +```{opcmd} show mpls table + +Show the MPLS table. +``` + +```{opcmd} show mpls pseudowire + +Show MPLS pseudowire interfaces. ``` ### Reset -```{opcmd} reset mpls ldp neighbor \<IPv4 or IPv6 address\> +```{opcmd} reset mpls ldp neighbor \<address\> -Use this command to reset an LDP neighbor/TCP session that is established +Reset the established LDP session with the specified neighbor. + +The `<address>` parameter accepts an IPv4 or IPv6 address. ``` -[wikipedia (mpls)]: <https://en.wikipedia.org/wiki/Multiprotocol_Label_Switching> +## Example + +The following configuration sets up a basic MPLS LDP-enabled router. +OSPF is used as the underlying IGP to provide reachability between LDP +peers' loopback addresses, which are then used as the LDP transport +addresses and router IDs. MPLS forwarding and LDP are enabled on the +interface connecting to the network. + +```none +set protocols ospf area 0 network '192.0.2.1/32' +set protocols ospf area 0 network '192.0.2.4/31' +set protocols ospf parameters router-id '192.0.2.1' +set protocols mpls interface 'eth1' +set protocols mpls ldp discovery transport-ipv4-address '192.0.2.1' +set protocols mpls ldp interface 'eth1' +set protocols mpls ldp interface 'lo' +set protocols mpls ldp router-id '192.0.2.1' +set interfaces ethernet eth1 address '192.0.2.5/31' +set interfaces loopback lo address '192.0.2.1/32' +``` + +```{note} +This example assumes a transit (P-router) role: MPLS forwarding is +enabled on the transit interface only, not on the loopback. If the +router needs to terminate MPLS-encapsulated traffic on its loopback +(PE-router role), also configure `set protocols mpls interface 'lo'`. +``` + +Apply the same configuration on the LDP peer (using its own loopback and +transit interface addresses) to establish a basic LDP session. |