Errata

This section lists outstanding issues with the documentation.

Platform and Infrastructure

• The list of JUNOS software releases required to support graceful Routing Engine switchover with 1-port Gigabit Ethernet PICs with SFP on various routing platform types also applies to 1-port Gigabit Ethernet PICs without SFP. [High Availability (for JUNOS 8.4), System Basics (for JUNOS 8.3 and earlier)]
• The table titled “Graceful Routing Engine Switchover PIC Support” in the JUNOS High Availability Configuration Guide includes incorrect release information for certain PICs. The correct information for the indicated PICs is as follows:

  Table 2: Graceful Routing Engine Switchover PIC Support

  PIC Type	M10i	M20	M40e	M120	M320	T320	T640	TX Matrix
  Adaptive Services PICs	7.3	7.3	7.3	–	–	–	–	–
  Adaptive Services II PICs	7.2	7.2	7.2	8.2	7.2	7.2	7.2	7.3
  Adaptive Services II PICs with Link Services IQ (LSQ) interfaces	7.6	7.6	7.6	8.2	7.6	7.6	7.6	n/a

  In addition, formation for the following PIC types needs to be added to the table:

  Table 3: Graceful Routing Engine Switchover PIC Support

  PIC Type	M10i	M20	M40e	M120	M320	T320	T640	TX Matrix
  Multiservices 100	8.1R2	–	8.1R2	8.2	8.1R2	8.1R2	8.1R2	8.1R2
  Multiservices 400	–	–	8.1R2	8.2	8.1R2	8.1R2	8.1R2	8.1R2
  Multiservices 500	–	–	–	8.3	8.3	8.3	8.3	8.3

• The documentation incorrectly states the following: “The JUNOS software saves the current core file (0) and the four previous core files, which are numbered 1 through 4 (from newest to oldest).” In reality, in the output of the show system core-dumps command after five or more core files are generated, the .0 file is the oldest while the .4 file is the newest and is overwritten each time a new core file is generated. [System Basics]
• The following note is omitted from the section titled “Requirements for Routers with a Backup Router Configuration” in Chapter 6 of the High Availability Configuration Guide.

  Note: If you have a backup router configuration in which multiple static routes point to a gateway from fxp0, you must configure prefixes that are more specific than the static routes or include the retain flag at the [edit routing-options static route] hierarchy level. For example, if you configure the static route 172.16.0.0/12 from fxp0 for management purposes, you must specify the backup router configuration as follows: backup-router 172.29.201.62 destination [172.16.0.0/13 172.16.128.0/13]

  [High Availability]

• The following text is omitted from the section titled “Configuring the Preshared Key for an IKE Policy” in Chapter 16 of the System Basics Configuration Guide.

  A local certificate is an alternative to the preshared key. Commit fails if either a preshared key or local certificate is not configured.

  [System Basics]

User Interface and Configuration

• J-Web Quick Configuration pages do not support IPv6 addressing and routing. [J-series Basic Configuration]
• The new bridge option to the show system statistics command, which displays system statistics on MX-series routers, is not documented. [System Basics and Services Command Reference]
• The new static-mac statement at the [edit routing-instances instance-name protocols l2vpn site site-name interface interface-name] and [edit routing-instances instance-name protocols vpls site site-name interface interface-name] is not documented. [VPNs]
• The name of the new ASP_L2TP_STATS_BULK_QUERY_FAILED system log tag appears in the documentation as ASP_L2TP_BULK_STATS_QUERY_FAIL, which is incorrect. Also, the actual logged text for the new tag is “Number of queries (number-queries) in statistics request exceeded limit (maximum-queries).” [System Log Messages]
• The name of the new ASP_SFW_ALG_LEVEL_ADJUSTED system log tag appears in the documentation as ASP_SFW_ADJUSTING_ALG_LEVEL, which is incorrect. [System Log Messages]

Interfaces and Chassis

• To display the FRU model number, part number, and serial number, issue the show chassis hardware models command. To display the FRU model number, part number, and CLEI code, issue the show chassis hardware clei-models command. [System Basics and Services Command Reference]
• FRF.12 is supported on link services (ls-) interfaces on the J-series routing platform. [Services Interfaces]
• The drop-and-insert multiplexer is now integrated into channelized T1/E1 PIMs on J-series Services Routers. The data-input (system | interface interface-name) statement at the [edit interfaces ds-pim/0/port:channel] hierarchy level is not documented. [Network Interfaces]
• VRRP for IPv6 can be configured for a global unicast address, not just for a site-local address. [Network Interfaces]
• In the chapter titled “System Management Configuration Statements,” the list of processes at the [edit system processes] hierarchy level shows the failover option for most processes. The correct syntax for this option is failover failover-option. Also, the correct syntax for the timeout statement at this level is timeout  seconds. [System Basics]

Services Applications

• The documentation does not include information about the clear-ike-sas-on-pic-restart statement, which you can include at the [edit services ipsec-vpn] hierarchy level to have the IKE SAs associated with tunnels cleared automatically when an AS PIC goes offline or is restarted. [Services Interfaces]

Layer 2 Ethernet Services

• Relationship between MAC address limit statements—The addresses that are added to the MAC address table can be learned at three source levels, and you can limit the number accepted from each level: from a single logical interface, from the interfaces in a bridge domain or VPLS instance, and from all interfaces on the entire router. The statements for setting the limits are as follows:

  • interface-mac-limit—Limits the number of addresses learned from a single logical interface. The default is 1024 addresses.

    For bridge domains, you can include the interface-mac-limit statement at the [edit bridge-domains bridge-domain-name bridge-options], [edit bridge-domains bridge-domain-name bridge-options interface interface-name], [edit routing-instances routing-instance-name bridge-domains bridge-domain-name bridge-options], or [edit routing-instances routing-instance-name bridge-domains bridge-domain-name bridge-options interface interface-name] hierarchy level.

    For VPLS, you can include the interface-mac-limit statement at the [edit routing-instances routing-instance-name protocols vpls], [edit routing-instances routing-instance-name protocols vpls site site-name interfaces interface-name], [edit logical-routers logical-router-name routing-instances routing-instance-name protocols vpls], or [edit logical-routers logical-router-name routing-instances routing-instance-name protocols vpls site site-name interfaces interface-name] hierarchy level.

  • mac-table-size—Limits the number of addresses learned from the interfaces in a domain or instance. The default is 5120 addresses.

    For bridge domains, you can include the mac-table-size statement at the [edit bridge-domains bridge-domain-name bridge-options] or [edit routing-instances routing-instance-name bridge-domains bridge-domain-name bridge-options] hierarchy level.

    For VPLS, you can include the mac-table-size statement at the [edit logical-routers logical-router-name routing-instances routing-instance-name protocols vpls] or [edit routing-instances routing-instance-name protocols vpls] hierarchy level.

  • global-mac-limit—Limits the number of addresses learned by all interfaces, domains, and instances configured on the router, and effectively sets the maximum size of the MAC address table. The default is 393,215 addresses. You can include the global-mac-limit statement at the [edit protocols l2-learning] hierarchy level.
  If you change one or more of these limits from its default, the largest limit for any interface in a domain or instance (set by the interface-mac-limit statement) must be lower than the limit for its parent domain or instance (set by the mac-table-size statement). Similarly, the largest limit for any domain or instance on the router must be lower than the global limit set by the global-mac-limit statement. If the configuration does not comply with this requirement, the commit operation fails. [Routing, VPNs, Feature Guide]

General Routing

• The manuals currently state that only the following routing platforms support GRES for VPLS: M10i, M20, M40e, M320, T320, and T640. The TX Matrix routing platform also supports GRES for VPLS. [System Basics, Feature Guide, VPNs]

Routing Protocols

• For the spanning tree (STP) bridge domain configuration and show spanning-tree bridge command shown in the JUNOS Routing Protocols Configuration Guide, the words “bridge domain” should read as “bridge.” In addition, the extended system ID is now configurable. [Routing]
• The learning-vlan-id option in the show bridge mac-table and show bridge mac-table interface commands has been changed to vlan-id. [Protocols Command Reference]
• The documentation for BFD with BGP, IS-IS, OSPF, and RIP incorrectly states that values for the bfd-liveness-detection minimum-interval statement smaller than 300 ms can cause BFD sessions to transition up and down. (You can include the statement at various hierarchy levels, including following: [edit protocols bgp], [edit protocols isis interface interface-name], [edit protocols ospf area area-id interface interface-name], and [edit protocols rip group group-name].) The smallest recommended value is now 100 ms for most routing platform types, but on some platform types even smaller values do not degrade performance. For more information, contact Juniper Networks customer support. [Routing]

MPLS Applications

• Inter-AS traffic engineering (Phase 2) enables traffic-engineered MPLS LSPs to dynamically discover OSPF autonomous system boundary routers (ASBRs) and enables routers to establish a traffic-engineered LSP across multiple autonomous systems (ASs). Each AS is assumed to be under the control of a single service provider and to use OSPF. To configure traffic engineering across multiple ASs using OSPF, include the traffic-engineering statement at the [edit protocols (ospf | ospf3) area area-id interface interface-name passive] hierarchy level. [MPLS Applications]
• The recommended range for static MPLS label-switched path (LSP) labels is now 1,000,000 through 1,048,575. The previous recommended range for static labels was 10,000 through 99,999. Configurations that have static labels outside the new range cannot be committed. To configure static MPLS labels, include the label-map statement at the [edit protocols mpls interface interface-name] hierarchy level. The JUNOS software now uses the range of 10,000 through 99,999 for processing label-switched interface (LSI) labels. [MPLS Applications]

VPNs

• When you configure VPLS without a Tunnel Services PIC by including the no-tunnel-services statement at the [edit routing-instances routing-instance-name protocols vpls] hierarchy level, the following limitations apply:
  • An Enhanced FPC is required.
  • Aggregated SONET/SDH interfaces used as core-facing interfaces are not supported.
  • Channelized interfaces used as core-facing interfaces are not supported.
  • ATM1 interfaces are not supported.

  [VPNs]

• For multiprotocol BGP-based multicast VPNs, existing vrf-import and vrf-export policies for importing and exporting VPN routes might prevent import or export of multicast VPN routes if the policies reject routes that match the protocol mvpn or combined protocol bgp and family inet-mvpn terms. The workaround is to relax the policies to accept routes that match these terms or to use multicast VPN-specific route targets to override the vrf-import and vrf-export policies. [Feature Guide, VPNs]
• When you configure a multicast VPN routing instance, the target value for a multicast VPN-specific route target must not be identical to a target value for a unicast route target configured in another routing instance. [VPNs]

Class of Service

• As stated in the documentation, you can selectively set the DSCP field of IPv4 and IPv6 packets to 0 without affecting output queue assignment, and continue to set the MPLS EXP field according to the configured rewrite table, based on forwarding classes. This feature is not supported with GRE and IP-IP tunnels. The documentation incorrectly implies that you can use the dscp 0 action modifier to set the DSCP field of IPv4 and IPv6 packets to 0. For IPv4 traffic, the dscp 0Service action modifier at the [edit firewall family inet filter filter-name term term-name then] hierarchy level is valid. However, for IPv6 traffic, you configure this feature by including the traffic-class 0 action modifier at the [edit firewall family inet6 filter filter-name term term-name then] hierarchy level. [CoS]
• On J-series Services Routers, if you create a policy to match IPv4 traffic with a route filter and assign the traffic to a forwarding class, then apply the policy at the [edit class-of-service forwarding-policy class policy-name] hierarchy level, use of the classification-override statement at the [edit class-of-service forwarding-policy class policy-name] hierarchy level is not supported. [J-series Advanced Configuration, CoS]
• For 4-port Fast Ethernet ePIMs on J-series Services Routers, if you apply a CoS scheduler map on outgoing (egress) traffic, the router does not divide the bandwidth appropriately among the CoS queues. As a workaround, configure enforced CoS shaping on the ports. [J-series Getting Started]

Routing Policy and Firewall Filters

• Floating point metric multiplier values in routing policies are now limited to eight significant digits. [Policy]

Network Management

• The documentation states incorrectly that the JUNOS software supports RFC 2737, Entity MIB (Version 2). [Network Management, Hierarchy and RFC Reference]

JUNOS XML API and Scripting

• In the HTML version of the JUNOS 8.4 Configuration and Diagnostic Automation Guide, line breaks are not preserved in most examples of scripts, SLAX code, and XSLT code, making it difficult to understand and use the samples. The formatting is correct in the PDF version. [Automation]