Changes in Default Behavior and Syntax in JUNOS Software Release 9.4 for M-series, MX-series, and T-series Routing Platforms

Hardware

• Combinations of PICs—On Juniper Networks routing platforms, you can typically install any combination of Physical Interface Cards (PICs) in a single Enhanced Flexible PIC Concentrator (FPC). Newer JUNOS services for some PICs can require significant Internet Processor ASIC memory, and some configuration rules might limit certain combinations of PICs on some platforms. To conserve memory, group PICs in the same family together on the same FPC. Ethernet and SONET/SDH PICs typically do not use large amounts of memory. Adaptive Services, Asynchronous Transfer Mode (ATM) 2, Gigabit Ethernet, and IQ serial PICs use more.

  Configuration rules might apply to PICs installed on standard Enhanced FPCs on the following routing platforms: M5, M10, M20, M40, M40e, M160, M320, J20, T320, and T640.

  Configuration rules do not apply to PICs installed in the following routers or FPCs:

  • J-series, M7i, M10i, or M120 routers
  • Enhanced Plus FPCs on M-series and J20 routers
  • Enhanced Scaling FPCs

  When you upgrade the JUNOS software, a warning appears if any configuration rules affect your PIC combinations. If you continue the installation, the PICs appear to be online (the LEDs are on), but the JUNOS software cannot enable them and they cannot pass traffic. As a workaround, you need to plan which PICs to install on the Enhanced FPCs or PIC slots on your routing platform. For specific information about PIC combination rules, consult Technical Bulletin PSN-2007-01-023. Go to http://www.juniper.net/customers/support and click Technical Bulletins. On the JTAC Technical Bulletins web page, enter PSN-2007-01-023 in the Search field, select the CS Technical Bulletin ID radio button, and click Search.

Platform and Infrastructure

• On MX-series routers, a maximum of 4000 active logical interfaces are now supported on a bridge domain or on each mesh group in a VPLS instance configured for Layer 2 bridging. [Routing Protocols Configuration Guide]

User Interface and Configuration

• Revised range for the ARP aging timer—The lower limit of the aging-timer statement at the [edit system arp] hierarchy level has been reduced to 1 minute. Therefore, the current range of this parameter is from 1 through 240 minutes. [System Basics]
• Configuration statements for disabling the reporting of ping record route and timestamp—Two new statements—no-ping-record-route and no-ping-time-stamp—have been introduced at the [edit system] hierarchy level. Include the no-ping-record-route statement in the configuration to prevent the Routing Engine from recording and displaying the route of the ping request packet in the response. Include the no-ping-time-stamp statement in the configuration to disable the Routing Engine from recording and displaying the timestamp in the ping response. By configuring these statements, you can prevent unauthorized users from discovering information about the PE router and its loopback address. [System Basics].
• New operational mode command show chassis network services (MX-series routers only)—To view the network services mode configured for the router, issue the show chassis network services command. The output displays either IP or Ethernet depending on the services mode (IP or Ethernet) configured on the router. [System Basics and Services Command Reference]

Interfaces and Chassis

• VRRP enhancements for ARP requests—When a router responds to an ARP request, the VRRP virtual MAC address is sent as the Ethernet source address in the Ethernet frame. When VRRP and proxy ARP are both configured, only the VRRP master on that subnet responds to the proxy ARP request. In previous software releases, when a router responded to an ARP request, the hardware MAC address was sent as the Ethernet source address. When VRRP and proxy ARP were both configured, the router responded as proxy for an ARP request if the address was reachable, irrespective of the VRRP state (backup or master) on the subnet. [Network Interfaces]
• New operational mode command show chassis network services (MX-series routers only)—To view the network services mode configured for the router, issue the show chassis network services command. The output displays either IP or Ethernet depending on the services mode (IP or Ethernet) configured on the router. [System Basics and Services Command Reference]

Routing Protocols

• Point-to point-interface support for OSPFv3—You can now configure an OSPFv3 interface as a point-to-point interface. Previously, only OSPFv2 interfaces could be configured as point-to-point interfaces. Include the interface-type p2p statement at the [edit protocols ospf3 area area-id interface interface-name] or [edit protocols ospf3 realm (ipv4-multicast | ipv4-unicast | ipv6-multicast) area area-id interface interface-name] hierarchy level. Use the show ospf3 interface command to display information about configured point-to-point interfaces. [Routing Protocols, CR: Routing Protocols]
• Traffic engineering shortcuts support extended to OSPFv3—Enables you to configure OSPFv3 to use IGP shortcuts for IPv6 routes. Previously, only IPv4 routes calculated through OSPFv2 routes were supported. Include the traffic-engineering shortcuts statement at the [edit protocols ospf3] hierarchy level. Label-switched paths (LSPs) to be used for shortcuts continue to be signaled using IPv4. However, by default, shortcut IPv6 routes calculated through OSPFv3 are added to the inet6.3 routing table. The default behavior is for BGP only to use LSPs in its calculations. If you configure MPLS so that both BGP and IGPs use LSPs for forwarding traffic, IPv6 shortcut routes calculated through OSPFv3 are added to the inet6.0 routing table. You can also configure OSPFv3 to ignore RSVP LSP metrics in traffic engineering shortcut calculations. Include the ignore-lsp-metrics statement at the [edit protocols ospf3 traffic-engineering shortcuts] hierarchy level. [Routing Protocols]
• Periodic Packet Management enabled by default on the Packet Forwarding Engine (M120, M320, MX-series, T-series, and TX-series routers)—Distributed Periodic Packet Management (PPM) is now enabled by default on the Packet Forwarding Engine. Previously, you had to enable PPM on the Packet Forwarding Engine. PPM also continues to be enabled by default on the Routing Engine. To disable PPM on the Packet Forwarding Engine, include the no-delegate-processing statement at the [edit routing-options ppm] hierarchy level. In addition, the delegate-processing statement at the [edit routing-options ppm] hierarchy level has been deprecated. This statement was previously used to enable PPM on the Packet Forwarding Engine. [Routing Protocols]
• With ce1 or e1 partitioned 4xCOC12/STM4 IQE and 1xCOC48/STM16 IQE PICs, the valid range for timeslots under e1-options in ce1 and e1 partitions is 2 through 32. This option is used to create fractional E1 interfaces. [Network Interfaces]

Routing Policy and Firewall Filters

• Prefix lists with IPv4 addresses for IPv6 firewall filters—You can now specify a prefix list that includes IPv4 addresses at the [edit firewall family inet6 filter filter-name term term-name from source-prefix-list prefix-list-name] or the [edit firewall family inet6 filter filter-name term term-name from prefix-list prefix-list-name] hierarchy level. Specify the name of a prefix list defined at the [edit policy-options prefix-list prefix-list-name] hierarchy level. The IPv4 addresses are ignored, and only the IPv6 addresses included in a prefix list specified in an IPv6 firewall filter are applied. Previously, the commit operation failed when a prefix list that included IPv4 addresses was specified in an IPv6 firewall filter. [Routing Policy]
• The output format of the show dhcp relay bindings detail command has changed from a tabular display to a line-by-line display. In addition, a new field, interface-name, was added to the output of this command. The interface-name field provides the MAC address of a client that is part of a DHCP relay/DHCP snooping configuration. [JUNOS Routing Protocols and Policies Command Reference]

VPNs

• Multiple Layer 2 circuit pseudowires and VPLS mesh groups—You can now terminate multiple Layer 2 circuit pseudowires at a single VPLS mesh group by including the local-switching statement in the mesh group configuration at the [edit routing-instances routing-instance-name protocols vpls mesh-group mesh-group-name] hierarchy level. Previously, you needed to configure a mesh group for each Layer 2 circuit pseudowire terminating at the router. [VPNs]
• The output of the show vpls connections extensive command now properly omits information related to inactive site IDs. Previously, when you changed a remote site ID, the output might have still referenced the original site ID. [Routing Protocols Command Reference]
• When configuring Draft-rosen multicast VPNs operating in source-specific mode and using the vrf-export statement, the export policy to be specified in the vrf-export statement must have a term that accepts routes from the vrf-name.mdt.0 routing table to ensure proper PE auto discovery with the inet-mdt family. This is done automatically when you configure the vrf-target statement to automatically generate the VRF's export policy. [VPNs, Multicast]

Forwarding and Sampling

• Configure load-balanced hash seed—On routing platforms with the Internet Processor II application-specific integrated circuit (ASIC), all PFE slots are assigned the same hash seed by default. A new per-flow hash-seed number statement is available which enables you to configure a unique, load-balanced hash seed for each PFE slot, enabling better utilization of available links.

  To configure, include the hash-seed number statement at the [edit forwarding-options load-balance per-flow] hierarchy level. The range you can configure for the hash-seed number is 0 through 65,535. The default is 0. [Routing Policy]

MPLS Applications

• LSP ping interval—You can now configure the time interval for the LSP ping command. Specify the LSP ping interval time using the lsp-ping-interval statement at the [edit protocols ldp oam] hierarchy level for LDP-signaled LSPs and at the [edit protocols mpls oam] hierarchy level for RSVP LSPs. The ping command for LDP-signaled LSPs is ping mpls ldp. The ping command for RSVP-signaled LSPs is ping mpls rsvp. [MPLS, System Basics Command Reference]

Forwarding and Sampling

• Configure load-balanced hash seed—On routing platforms with the Internet Processor II application-specific integrated circuit (ASIC), all PFE slots are assigned the same hash seed by default. A new per-flow hash-seed number statement is available which enables you to configure a unique, load-balanced hash seed for each PFE slot, enabling better utilization of available links.

  To configure, include the hash-seed number statement at the [edit forwarding-options load-balance per-flow] hierarchy level. The range you can configure for the hash-seed number is 0 through 65,535. The default is 0. [Routing Policy]

Related Topics

• Features in JUNOS Software Release 9.4 for M-series, MX-series, and T-series Routing Platforms
• Issues in JUNOS Software Release 9.4 for M-series, MX-series, and T-series Routing Platforms
• Errata and Changes in Documentation for JUNOS Software Release 9.4 for M-series, MX-series, and T-series Routing Platforms
• Upgrade and Downgrade Instructions for JUNOS Software Release 9.4 for M-series, MX-series, and T-series Routing Platforms