traffic-engineering (Protocols IS-IS)

Syntax

Hierarchy Level

Description

Configure traffic engineering properties for IS-IS.

IS-IS always performs shortest-path-first (SPF) calculations to determine next hops. For prefixes reachable through a particular next hop, IS-IS places that next hop for that prefix in the inet.0 routing table. In addition, for routers running MPLS, IS-IS installs the prefix for IPv4 routes in the inet.3 routing table as well. The inet.3 table, which is present on the ingress router, contains the host address of each MPLS label-switched path (LSP) egress router. BGP uses this routing table to resolve next-hop addresses.

If you enable IS-IS traffic engineering shortcuts and if there is a label-switched path to a point along the path to that prefix, IS-IS installs the prefix in the inet.3 routing table and uses the LSP as a next hop. The net result is that for BGP egress routers for which there is no LSP, BGP automatically uses an LSP along the path to reach the egress router.

In Junos OS Release 9.3 and later, IS-IS traffic engineering shortcuts support IPv6 routes. LSPs to be used for shortcuts continue to be signaled using IPv4. However, by default, shortcut routes calculated through IPv6 routes are added to the inet6.3 routing table. The default behavior is for only BGP to use LSPs in its calculations. If you configure MPLS so that both BGP and interior gateway protocols use LSPs for forwarding traffic, shortcut routes calculated through IPv6 are added to the inet6.0 routing table. IS-IS ensures that the IPv6 routes running over the IPv4 MPLS LSP are correctly de-encapsulated at the tunnel egress by pushing an extra IPv6 explicit null label between the IPv6 payload and the IPv4 transport label.

RSVP LSPs with a higher preference than IS-IS routes are not considered during the computation of traffic engineering shortcuts.

To configure IS-IS so that it uses LSPs as shortcuts when installing information in the inet.3 or inet6.3 routing table, include the following statements:

For IPv4 traffic, include the inet statement. For IPv6 traffic, include the inet6 statement.

To configure IPv4 MPLS or IPv6 MPLS shortcuts explicitly for segment routing, include the inet-mpls statement for IPv4 MPLS traffic and the inet6-mpls statement for IPv6 MPLS traffic.

To configure load balancing across multiple LSPs, include the multipath statement.

When traffic engineering shortcuts are used, RSVP first looks at the metric2 value, which is derived from the IGP cost. After this, RSVP considers the LSP metric value. So, if a certain path changes for an LSP and the cost changes, not all LSPs are used to load- balance the network.

When a route with an improved metric is added to the IS-IS internal routing table, IS-IS flushes all next-hop information (including LSP next-hop information) for a route. This is undesirable, because certain equal-cost multipath (ECMP) combinations can be lost during route calculation. To override this default behavior for load balancing, include the lsp-equal-cost statement to retain the equal cost path information in the routing table.

Because the inet.3 routing table is present only on ingress routers, you can configure LSP shortcuts only on these routers.

Default

IS-IS traffic engineering support is enabled.

By default, IS-IS supports traffic engineering by exchanging basic information with the traffic engineering database. To disable this support, and to disable IS-IS shortcuts if they are configured, include the disable statement.

Options

credibility-protocol-preference

Specify that IS-IS should use the configured protocol preference for IGP routes to determine the traffic engineering database credibility value. By default, the traffic engineering database prefers IS-IS routes even when the routes of another IGP are configured with a lower, that is, more preferred value. Use this statement to override this default behavior.

The traffic engineering database assigns a credibility value to each IGP and prefers the routes of the IGP with the highest credibility value. In Junos OS Release 9.4 and later, you can configure IS-IS to take protocol preference into account to determine the traffic engineering database credibility value. When protocol preference is used to determine the credibility value, IS-IS routes are not automatically preferred by the traffic engineering database, depending on your configuration. For example, OSPF routes have a default preference value of 10, whereas IS-IS Level 1 routes have a default preference value of 15. When protocol preference is enabled, the credibility value is determined by deducting the protocol preference value from a base value of 512. Using default protocol preference values, OSPF has a credibility value of 502, whereas IS-IS has a credibility value of 497. Because the traffic engineering database prefers IGP routes with the highest credibility value, OSPF routes are now preferred.

Note:

This feature is also supported for OSPFv2.

disable

At the [edit protocols isis traffic-engineering] hierarchy level, disable IS-IS support for traffic engineering. By default, IS-IS support for traffic engineering is enabled.

ignore-lsp-metrics

Ignore the metrics for RSVP label-switched paths (LSPs) in IS-IS traffic engineering shortcut calculations or when you configure LDP over RSVP LSPs. If you are using RSVP for traffic engineering, you can run LDP simultaneously to eliminate the distribution of external routes in the core. The LSPs established by LDP are tunneled through the LSPs established by RSVP. LDP effectively treats the traffic-engineered LSPs as single hops. Ignoring the metric of RSVP LSPs avoids mutual dependency between IS-IS and RSVP, eliminating the time period when the RSVP metric used for tunneling traffic is not up to date.

ipv6

Configure IS-IS traffic engineering to store IPv6 information in the traffic engineering database (TED) in addition to IPv4 addresses. BGP-LS distributes this information as routes from the TED to the lsdist.0 routing table using the TED import policies. These routes are advertised to BGP-TE peers as network layer reachability information (NLRI) with IPv6 router ID type, length, and value (TLV).

multipath

Enable load balancing for multiple LSPs.

lsp-equal-cost

Configure LSPs to be retained as equal cost paths for load balancing when a better route metric is added to the routing table.

When a route with an improved metric is added to the IS-IS internal routing table, IS-IS flushes all next-hop information (including LSP next-hop information) for a route. This is undesirable, because certain equal-cost multipath (ECMP) path combinations can be lost during route calculation. To override this default IS-IS behavior, include the lsp-equal-cost statement for load balancing, so that the equal cost path information is retained in the routing table.

l3-unicast-topology

Download IGP topology information into the traffic engineering database (TED). In Junos OS, the IGPs install topology information into a database called the traffic engineering database. The traffic engineering database contains the aggregated topology information. The IGP routes are installed by the traffic engineering database on behalf of the corresponding IGP into a user-visible routing table called lsdist.0, subject to route policies.

Note:

In earlier Junos releases the l3-unicast-topology option was referred to as igp-topology. The igp-topology syntax is deprecated.

shortcuts

Configure IS-IS to use MPLS label-switched paths (LSPs) as next hops if possible when installing routing information into the inet.3 or inet6.3 routing table. Internal gateway protocol (IGP) shortcuts allow the IGP to install prefixes in inet.3 or inet6.3. It is only necessary to enable IGP shortcuts on the ingress router because that is the router performing the shortest-path-first (SPF) calculations.

It is important to understand how IGP shortcuts affect the protocol and routing table relationship. The IGP performs SPF calculations to subnets downstream of LSP egress points, but the results of these calculations are entered into the inet.3 table only. At the same time, the IGP performs its traditional SPF calculations and enters the results of these calculations into the inet.0 table. The result is that although the IGP is making entries into the inet.3 table, BGP is still the only protocol with visibility into that table for the purposes of route resolution. Therefore, forwarding to AS-internal destinations still uses the inet.0 IGP routes, and the LSPs are only used for BGP next-hop resolution. If you want the LSPs to be used for IGP next-hop resolution, you must configure traffic-engineering bgp-igp.

multicast-rpf-routes

Install unicast IPv4 routes into the multicast routing table (inet.2) for multicast reverse-path-forwarding (RPF) checks. Traffic engineering shortcuts must be enabled. IPv4 multicast topology must not be enabled. Label-switched paths (LSPs) must not be advertised into IS-IS.

The remaining statements are explained separately. See CLI Explorer.

Required Privilege Level

routing—To view this statement in the configuration.

routing-control—To add this statement to the configuration.

Release Information

Statement introduced before Junos OS Release 7.4.

Support for the family statement introduced in Junos OS Release 9.3.

Support for the credibility-protocol-preference statement introduced in Junos OS Release 9.4.

Support for the multipath statement introduced in Junos OS Release 9.6.

Support for the lsp-equal-cost statement introduced in Junos OS Release 9.6.

Support for inet-mpls and inet6-mpls statements introduced in Junos OS Release 17.2 for MX Series, PTX Series, QFX5100 switches, and QFX10000 line switches.

Support for inet-mpls and inet6-mpls statements introduced in Junos OS Release 17.3 for QFX5110 and QFX5200 switches.

Support for igp-topology statement introduced in Junos OS Release 17.4R1 for MX Series, and PTX Series. The igp-topology syntax is now deprecated. Current Junos releases use the l3-unicast-topology statement.

ipv6 option introduced in Junos OS Release 20.4R1.

ON THIS PAGE