Previous Releases

Release 9.0R3

The following issues have been resolved since JUNOS Release 9.0R3.6. The identifier following the description is the tracking number in our bug database.

Software Installation

• When you issue the request system partition hard-disk command, the hard disk repartition fails and the disk becomes unusable. The disk can be recovered by taking a snapshot from the compact flash card and rebooting the router. [PR/269493: This issue has been resolved.]

Platform and Infrastructure

• When you issue the show system firmware command, the Current version field does not show the correct version of the service PICs. [PR/64918: This issue has been resolved.]
• If both TCP and UDP probe servers are configured (the tcp and udp statements are both included at the [edit services rpm probe-server] hierarchy level) and another router sends a probe of each type simultaneously, multiple occurrences of the RMOPD_SENDMSG_FAILURE message are written to the system log. [PR/66570: This issue has been resolved.]
• During a Routing Engine switchover, Flexible PIC Concentrators (FPCs) might reset multiple times. [PR/70857: This issue has been resolved.]
• On M7i and M10i routers, when the system log for the CFEB becomes full, additional messages are discarded instead of overwriting the oldest messages in the log. [PR/79128: This issue has been resolved.]
• On T-series routing platforms, if you include the indirect-next-hop statement at the [edit routing-options forwarding-table] hierarchy level for VPN routes, routing ASIC SRAM utilization increases by approximately 30 percent, or 8 bytes per route. [PR/98738: This issue has been resolved.]
• The command show chassis routing-engine may display incorrect information for CPU temperature, CPU utilization, and Load averages in some releases. [PR/99728: This issue has been resolved.]
• When graceful Routing Engine switchover (GRES) and multicast are both configured on a router, the master Routing Engine kernel might dump core because of inconsistencies between the multicast forwarding database on the master Routing Engine and the multicast forwarding database on the backup Routing Engine. [PR/100795: This issue has been resolved.]
• When using aggregate bundles with FRR, it takes about 10 to 16 seconds packet loss when one of the member links fail. [PR/101295: This issue has been resolved.]
• The Packet Forwarding Engine might incorrectly log an error like the following when a member link in an aggregate goes down or up: “RT_PFE: NH IPC op 34 (GET NH BULK STATS) failed, err 5 (Invalid)." [PR/105841: This issue has been resolved.]
• When you enable point-to-multipoint LSPs over an outgoing aggregated Ethernet interface that is configured with circuit cross-connect (CCC) switching, the LSP fails to forward traffic and the following error appears in the system log: "nh_ucast_add." As a workaround, disable the interface and LSP, reenable them in that order, and then clear the RSVP session for the LSP. [PR/105884: This issue has been resolved.]
• In certain rare circumstances, an M-series router might generate a core file and restart when it tries to determine the best match for a specified route prefix. [PR/239837: This issue has been resolved.]
• If Layer 2 encapsulation is larger than 34 bytes, the Packet Forwarding Engine might restart unexpectedly. [PR/240080: This issue has been resolved.]
• When the management interface (fxp0) initialization does not complete, the interface loses network connectivity and does not respond to any commands within the timeout period of 10 milliseconds. [PR/253479: This issue has been resolved.]
• Using the -ox option with the smartd process is not recommended on mounted devices because it may result in unexpected behavior. [PR/255473: This issue has been resolved.]
• On an MX platform, if the first interface has family MPLS configured, a DPC with a 10x1 Gigabit Ethernet (GE) interface might report the same interface statistics on all ports of the same PIC slot . As a workaround don't have a family interface configured on the first port of the PIC slot. [PR/262607: This issue has been resolved.]
• When you change the interface configuration from point-to-point encapsulation to Frame Relay encapsulation, the operating system might generate a core file and stop operating. [PR/265025: This issue has been resolved.]
• When IPv6 traffic is present on MLPPP bundles using MultiServices PIC, the Service PIC may core due to excessive logging as a result of invalid checksum calculation. [PR/266214: This issue has been resolved.]
• When an FPC is brought online immediately after being taken offline, it may not be properly initialized due to a race condition in timing between the FPC and Routing Engine. [PR/272086: This issue has been resolved.]
• SNMP statistics might incorrectly decrease on an interface. [PR/274553: This issue has been resolved.]
• On a T1600 router, the Link Aggregation Control Protocol (LACP) does not work on the following Ethernet interfaces: 10x1 Gigabit Ethernet, IQ2, or Fast Ethernet. [PR/274586: This issue has been resolved.]
• On MX-series routers, when unicast RPF is configured on an interface (the rpf-check statement is included at the [edit interfaces interface-name unit logical-unit-number family inet] hierarchy level), the DPC that houses the interface might generate a core file. [PR/275466: This issue has been resolved.]
• When using set forwarding-options sampling output file filename sampled.pkts, the sampled.pkts file (which should be created in /var/tmp in the router) is not being created. If restart sampling is called, the file gets created. This is because the j-flow license database is not working properly when sampling is configured. A workaround for this is to issue the show system license usage command. The license database is then updated and the file is created. [PR/275473: This issue has been resolved.]
• On a J-series platform running Release 8.5R2, the value displayed for idle CPU utilization is not correct. This is only a display issue, and does not affect router operation. [PR/275541: This issue has been resolved.]
• When the multipath statement is configured at the [edit routing-instances] hierarchy level and the indirect-next-hop statement is configured at the [edit routing-options forwarding-table] hierarchy level, the following error message might be displayed: “fpc6 NH: Unilist nh (1058732) failed addition to jtree (9).” The error indicates that there might be some packet loss because the next hop referenced by the mulitpath prefixes might not be installed. [PR/275586: This issue has been resolved.]
• When a GGSN C-PIC sends a packet larger than the MTU of the outgoing interface in a default VRF, ICMP error messages that indicate fragmentation is needed do not reach the C-PIC. [PR/276392: This issue has been resolved.]
• If an aggregated Ethernet or aggregated SONET bundle has a large number of aggregate next hops, when a new child link is added or a child interface goes down and comes up, the Packet Forwarding Engine might generate a core file. [PR/276424: This issue has been resolved.]
• If the Packet Forwarding Engine restarts frequently, it might generate a core file. [PR/276539: This issue has been resolved.]
• When packets are queued for several seconds because of interface congestion, packet CRC errors are reported in some cases. In other cases, the egress interface stops forwarding traffic (either all traffic is halted or only packets larger than 320 bytes are not forwarded). As a workaround, configure queues with a transmit rate of at least one percent of the line rate. When using strict-high priority queues, include policers to prevent interface congestion and the starving of lower-priority queues. Another alternative is to use high priority queues instead of strict-high priority queues. [PR/277853: This issue has been resolved.]
• Once protocol tracing is enabled, IO writes to the hard drives might get blocked and daemons sending packets will fall behind. This applies only to JUNOS 8.5R2 or higher. [PR/278580: This issue has been resolved.]
• When graceful Routing Engine switchover (GRES) is configured on a dual Routing Engine system, the backup Routing Engine might generate a core file. [PR/278901: This issue has been resolved.]
• On a J-series router, when you issue the set chassis fpc 0 pic 0 mlfr-uni-nni-bundles command, FPC 0 might go offline. [PR/279348: This issue has been resolved.]
• Swapping an IQ2 PIC with a Services PIC in the same PIC slot may cause the router to crash. [PR/280505: This issue has been resolved.]
• There is an issue with converting floating point format to ASCII format. [PR/279405: This issue has been resolved.]
• PIC based sampling (sp=-0/0/0) is not working. [PR/280766: This issue has been resolved.]
• On M10i routers that have Channelized DS3 IQ PICs installed, the Compact Forwarding Engine Board (CFEB) might generate a core file, which also interrupts FPC operation. [PR/283943: This issue has been resolved.]
• When you issue the request system software add command and include a file that does not have the .tgz extension, even though it may be identical to the file with the .tgz extension available from the Juniper Networks Support Web site, the router reboots. To avoid this problem, use the filenames available from the Juniper Networks Support Web site. [PR/283948: This issue has been resolved.]
• Under certain circumstances, DHCP discover packets might be leaked to all configured VRFs. [PR/286139: This issue has been resolved.]
• In JUNOS Release 8.5 and later, the Routing Engine might perform a reset when the hard drive performs a standby request. This problem is due to a watchdog error and does not generate a core file. For a possible workaround, contact the Juniper Networks Technical Assistance Center (JTAC). [PR/288011: This issue has been resolved.]
• By default, the Layer 2 address learning process (l2ald) turns on debug tracing without a traceoptions configuration, which might slow down the route update process to the Packet Forwarding Engine. [PR/290540: This issue has been resolved.]
• When a packet larger than the IP MTU size is transmitted, it registers as a microcode error rather than an MTU error. [PR/294485: This issue has been resolved.]
• On routers configured with aggregated SONET or aggregated Ethernet interfaces and multicast next hops, when the aggregated interface flaps, the kernel might restart unexpectedly. [PR/298073: This issue has been resolved.]
• On platforms with dual Routing Engines, the Routing Engines might dump core during processing of a BGP UPDATE message with a NEXT_HOP attribute that is a broadcast address of a local interface. [PR/302236: This issue has been resolved.]

User Interface and Configuration

• After upgrading to JUNOS Release 8.5R1.13, the message of the day is displayed twice. [PR/268625: This issue has been resolved.]
• When you issue the copy command for configuration at the [edit groups] hierarchy level, the management process (mgd) exits and generates a core file. [PR/269034: This issue has been resolved.]
• When a large number of router-filter entries (in the order of thousands) are configured at the [edit policy-options policy-statement policy-name term term-name] hierarchy level, it might take a few minutes to complete commands such as rollback number and show system rollback number compare number. [PR/272350: This issue has been resolved.]
• When you add the source-address statement at the [edit system syslog] hierarchy level and commit the configuration, the JUNOS system logging utility continues to use the previous source address. [PR/272434: This issue has been resolved.]
• JUNOS devices cannot be managed by Session and Resource Control (SRC) software. [PR/273117: This issue has been resolved.]
• When two users have a telnet or SSH session on a router, one in configure private mode and the other in configure mode, the telnet session disconnects if the user in configure mode issues the load patch command. [PR/274372: This issue has been resolved.]
• When you issue the replace pattern command, the telnet session might terminate and the management (mgd) process might generate a core file. [PR/274830: This issue has been resolved.]
• A JUNOS-ES device that is in in-device policy mode might allow rule creation for an IDP exempt rule that omits the required rule name. [PR/275418: This issue has been resolved.]
• A user without the maintenance access privilege level cannot issue the su root command from the shell to become a superuser. [PR/277888: This issue has been resolved.]
• Even though the trace permission is included at the [edit system login class class-name permissions] hierarchy level, users who belong to the login class receive the following error when they issue the show log command: "error: permission denied: log." As a workaround, add the "trace-admin" permission to the list of permissions. [PR/278950: This issue has been resolved.]
• When the filename in the event-script statement is not included at the [edit event-options policy policy-name then] hierarchy level, the event policy process (eventd) might generate a core file. [PR/290515: This issue has been resolved.]
• When a configuration group containing a wildcard match for a static route and qualified next hop of a broadcast interface is applied, the routing protocol process (rpd) might exit and dump core. [PR/290712: This issue has been resolved.]
• When you configure the ip-address at the [edit system radius-options attributes nas-ip-address] hierarchy level in JUNOS Release 8.5 and later, the nas-ip-address attribute is not included in the RADIUS packets. [PR/292274: This issue has been resolved.]

Interfaces and Chassis

• For aggregated Ethernet interfaces on T640 and TX Matrix platforms, the show interfaces extensive command sometimes reports extremely large incorrect values in the Dropped packets column of the Queue counters output. As a workaround, issue the clear interfaces statistics command. [PR/65857: This issue has been resolved.]
• When you commit firewall and RPF configurations, an erroneous "nh_jtree_fe_prehandler" message might appear on the Packet Forwarding Engine. This message is informational only and does not indicate an error condition. [PR/96146: This issue has been resolved.]
• On a dual Routing Engine system with graceful Routing Engine switchover (GRES) enabled, when an IPv6 interface is configured with the loopback statement at the [edit interfaces interface-name gigether-options] hierarchy level, the backup Routing Engine might report kernel replication errors in the output of the show system switchover command. [PR/102164: This issue has been resolved.]
• When graceful Routing Engine switchover (GRES) is enabled and you use the request system reboot command to reboot the master Routing Engine, if an FPC does not establish its connection to the new master Routing Engine before the previous master shuts down, the FPC restarts. [PR/234207: This issue has been resolved.]
• When graceful Routing Engine switchover (GRES) is enabled and the backup Routing Engine is taken offline, a CHASSISD_SNMP_TRAP message might not be generated. [PR/238797: This issue has been resolved.]
• Very rarely when there is a PIC error and the PIC is coming online again, the system might reset unexpectedly. With the fix, the PIC fails to attach if it has a hardware error. It does not cause the FPC to reset as in the previous case. [PR/241092: This issue has been resolved.]
• On a J-series router, when you upgrade a serial interface to JUNOS Release 8.0 and later, Frame Relay encapsulation might not work. Frame Relay does work with JUNOS Releases 7.0 though 7.6. [PR/241610: This issue has been resolved.]
• Memory might not be initialized correctly on some types of ASIC used on DPCs. [PR/255204: This issue has been resolved.]
• When you configure an IPv6 address as the primary or preferred address for an interface (by including the primary or preferred statement at the [edit interfaces interface-name unit logical-unit-number family inet6 address ipv6-address] hierarchy level) and commit the configuration, messages like the following are written to the system log: "DCD_CONFIG_WRITE_FAILED: Interface 'interface-name' configuration write failed for an IFA CHANGE: Operation not supported." [PR/258531: This issue has been resolved.]
• When you configure MLPPP or MLFR UNI NNI (FRF.16) bundles on link services IQ interfaces, a certain mix of traffic might cause a lower-priority queue to be starved when packets expire after not being scheduled for some time. [PR/262901: This issue has been resolved.]
• When a Fast Ethernet interface is connected to a Gigabit Ethernet interface that is configured for full duplex without autonegotiation, the information for the Fast Ethernet interface is incorrect in the "Autonegotiation information" section of the output from the show interfaces extensive command. [PR/263957: This issue has been resolved.]
• Under loaded conditions, the show interfaces rlsq command output might display incorrect statistical information because the statistics replies did not arrive in time. [PR/270467: This issue has been resolved.]
• When Routing Engine mastership is repeatedly switched, routing information maintained on the master and backup Routing Engine might be out of sync, causing all Packet Forwarding Engines to reset. [PR/271141: This issue has been resolved.]
• When you issue the show interfaces extensive command for an interface to which a Layer 2 input or output policer is applied, the value in the Dropped frames field for the policer might be a negative number. [PR/272971: This issue has been resolved.]
• When graceful Routing Engine switchover (GRES) and LSQ (rlsq) interfaces are configured, the last change field in the output of the show interface redundancy command might be incorrect after a Routing Engine switchover. [PR/273248: This issue has been resolved.]
• LSQ redundancy configured with graceful Routing Engine switchover (GRES) does not operate correctly when you add or delete a member T1 or E1 link after GRES switchover. As a workaround, deactivate and reactivate the entire LSQ (rlsq) bundle when individual links are added or deleted after GRES. [PR/273528: This issue has been resolved.]
• On the M320, a signal integrity issue in old clocking hardware might generate inaccurate alarms and errors when the actual clock is working perfectly. This behavior has no operational impact and has been fixed in later releases. [PR/275308: This issue has been resolved.]
• When you restart an IQ2 PIC, interfaces might not be taken down correctly. As a result, traffic continues to be directed to the interfaces and is lost. [PR/276125: This issue has been resolved.]
• For interfaces in an Enhanced III FPC in an M320 router, when you apply a firewall filter that includes the next term statement at the [edit firewall filter filter-name then] hierarchy level, the filter rejects all incoming traffic. [PR/278325: This issue has been resolved.]
• For a routing node in a routing matrix, when you remove a hardware component from the chassis, alarms are cleared for that component (which is correct), and also for all components of the same type. [PR/278672: This issue has been resolved.]
• On MX-series routers, if an interface in a VPLS routing instance is configured as nontagged (the encapsulation ethernet-vpls statement is included at the [edit interfaces interface-name] hierarchy level but the vlan-tagging statement is not), the interface does not forward frames with dual tags. [PR/279669: This issue has been resolved.]
• When checking for duplicate routes, the route mask of a particular route is compared with itself, resulting in failure of the Virtual Router Redundancy Protocol (VRRP) track route configuration to commit. [PR/280430: This issue has been resolved.]
• If you power off and power on a Routing Engine model RE-A-2000 on a T640 routing node (by issuing the request system power-off other-routing-engine and request system power-on other-routing-engine commands), the output of the show chassis hardware command no longer includes an entry for SPMB 1. [PR/281463: This issue has been resolved.]
• When a physical interface in a link services IQ multilink bundle goes down and comes up again, the PIC might generate a core file and write a message like the following to the PIC console: "coredump requested by CPU 12: thread hogged CPU." As a workaround, include the drop-timeout 0 statement at the [edit interfaces lsq-fpc/pic/0] hierarchy level to prevent the failure. [PR/281696: This issue has been resolved.]
• Adding per-unit-scheduler configuration to a one- or two-port IQ PIC might cause errors and affect the forwarding state of the ports. [PR/282934: This issue has been resolved.]
• XGE PICs on M120 routers take an unusually long time (up to 1.5 seconds) to send remote-fault messages. [PR/287147: This issue has been resolved.]
• Prior to JUNOS Releases 9.0R3 and 9.1R2, you cannot unplug a 4-port 10–Gigabit Ethernet PIC without taking the PIC offline first. [PR/288077: This issue has been resolved.]
• When you insert an OC192 SONET/SDH PIC that uses XFP optics into an Enhanced Type 3 FPC on a T640 routing node, the FPC might generate a core file. [PR/288884: This issue has been resolved.]
• Under the following conditions, a logical interface configured for VRRP (the vrrp-group statement is included at the [edit interfaces interface-name unit logical-unit-number family family address address] hierarchy level) does not initialize properly and the output for it from the show vrrp summary command displays the value bringup in the VR State field: (1) the logical interface is configured with dual VLAN tags (the vlan-tags statement is included at the [edit interfaces interface-name unit logical-unit-number] hierarchy level, (2) the configuration for another logical interface of the same physical interface includes the vlan-id statement at the [edit interfaces interface-name unit logical-unit-number] hierarchy level. The problem can occur even though the interfaces do not belong to the same VRRP group. [PR/288975: This issue has been resolved.]
• When multicast packets are replicated to multiple outbound interfaces at a moderate traffic load, transmit packets might be corrupted. There is no workaround. [PR/289353: This issue has been resolved.]
• On MX-series routing platforms, if you take fabric planes offline and the spare planes become active, you might see high traffic drops or continuous high fabric red drops. To recover from continuous high fabric drops, you must switch the fabric planes again. [PR/291541: This issue has been resolved.]
• SONET interfaces which are configured with interface hold up/down timers may remain down after an FPC reset or a PIC reset. To restore the interface, (temporarily) remove the interface hold timers. [PR/291707: This issue has been resolved.]
• On routers running JUNOS Release 9.0R1 or later, if VPLS multihoming is enabled and one instance has two or more sites configured with aggregated interfaces, it might trigger an MPLS loop, because the CCC-down flag for the interface is not applied and the router keeps learning MAC addresses on this aggregated interface. [PR/295634: This issue has been resolved.]

Services Applications

• If you configure a large number of TCP flows (for example, 500,000) and continuously shift the traffic between two Adaptive Services or MultiServices PICs, the PICs might generate core files. [PR/235646: This issue has been resolved.]
• When you commit a configuration that does not include either the pre-shared-key statement or the local-certificate statement at the [edit security ike policy policy-name] hierarchy level, the key management process (kmd) generates a core file. [PR/267957: This issue has been resolved.]
• Services PICs (such as the Adaptive Services and MultiServices PICs) do not record correct information in the SAMPLE-RATE field in the header of the cflowd packets that they export. [PR/276142: This issue has been resolved.]
• In a scaling setup with 4000 MLPPP sessions and traffic consisting of 256-byte packets, after traffic flows for more than an hour, the L2TP process does not start when the restart l2tp-service command is issued. [PR/276255: This issue has been resolved.]
• A Multiservices PIC running cRTP may crash and dump core when a large number of contexts are established. [PR/277000: This issue has been resolved.]
• The RTSP ALG implementation is not compatible with some RTSP server implementations. [PR/292961: This issue has been resolved.]
• The router handles data paths for Layer 2 Tunneling Protocol (L2TP) tunneled Multilink Point-to-Point Protocol (MLPPP) packets with a payload size of zero bytes incorrectly, which causes the service PIC to reset unexpectedly. [PR/296390: This issue has been resolved.]

General Routing

• When specifying the value of the members statement at the [edit policy-options community community-name] hierarchy level, if you use a regular expression that includes parentheses, the JUNOS software accepts values in which the parentheses are not balanced. As a result, the community might include unintended members or not include intended ones. [PR/104683: This issue has been resolved.]
• When the value of the members statement at the [edit policy-options community community-name] hierarchy level is a regular expression that begins with an integer, the JUNOS software does not always determine correctly whether the expression represents regular communities or extended communities. [PR/251510: This issue has been resolved.]

Routing Protocols

• The metric-out statement at the [edit protocols protocol-name group] hierarchy level incorrectly takes precedence over the metric-out statement configured under the neighbor configuration for the same group. [PR/31848: This issue has been resolved.]
• BGP traceoptions incorrectly reports Path Attribute flags with the EXT bit always reset. [PR/51953: This issue has been resolved.]
• When routes are propagated across IBGP, the show bgp group statistics command output does not display AS numbers correctly. [PR/69098: This issue has been resolved.]
• If more than 1000 communities are attached to a route, the routing process (rpd) might become unresponsive. Removing the communities and restarting the routing process might be necessary to recover. [PR/77001: This issue has been resolved.]
• If BGP route-target filtering is configured on the router (the route-target statement is included at the [edit protocols bgp family] hierarchy level), the routing protocols process (rpd) might generate a core file during initialization. [PR/77495: This issue has been resolved.]
• When a new receiver joins a multicast group that already has a receiver upstream of the RP, the new receiver might not receive multicast traffic for up to 120 seconds. [PR/228708: This issue has been resolved.]
• When PIM receives an (*,G) leave and an (S,G) join for an (S,G) entry that has been pruned and when the neighbor entry is deleted, the JP state for the (S,G) join is not deleted, resulting in PIM going into an infinite loop. [PR/235978: This issue has been resolved.]
• When unicast reverse path forwarding is configured (the rpf-check statement is included at the [edit interfaces interface-name unit logical-unit-number family family-name] hierarchy level), the lists of routes maintained by the routing protocols process (rpd) and the operating system might become different. [PR/252489: This issue has been resolved.]
• The output from the show route advertising-protocol bgp neighbor-address community community-id command is not correct if you specify a particular value (such as 11111:2222) for community-id. As a workaround, specify the wildcard value *.* instead. [PR/265624: This issue has been resolved.]
• When you activate or deactivate an aggregate route filter (represented by the aggregate statement at the [edit routing-options rib routing-table] hierarchy level), its contributing members are not reevaluated and the filter continues to function as before the change. [PR/270115: This issue has been resolved.]
• When you change the route distinguisher associated with a virtual private LAN service (VPLS) routing instance, the routing protocol process (rpd) might stop. [PR/270204: This issue has been resolved.]
• The show ospf route detail command output displays optional-capability value for intra-area router routes only. [PR/273809: This issue has been resolved.]
• IS-IS hello packets might not be generated for 8 to 12 seconds during nonstop active routing (NSR) switchover when the Periodic Packet Management process (ppmd) is not preprogrammed ahead of the switchover time. [PR/276823: This issue has been resolved.]
• When both of the following conditions apply, a change in interface status (up or down) causes a BGP status change: (a) there are more than 255 unnumbered interfaces without a destination address (the unnumbered-address lo0.0 statement is included at the [edit interfaces interface-name unit logical-unit-number family inet] hierarchy level for more than 255 logical interfaces), and (b) the BGP local address (specified by the local-address statement at the [edit protocols bgp group group-name] hierarchy level) is the last one in the list of addresses included at the [edit interfaces lo0 unit 0 family inet] hierarchy level. As a workaround, either use an unnumbered interface that has a destination address or do not set the BGP local address to an unnumbered interface. [PR/277202: This issue has been resolved.]
• The Multicast Source Discovery Protocol (MSDP) incorrectly reports a nonexistant security association (SA), resulting in the SA remaining in Protocol Independent Multicast (PIM) when it is deleted in the MSDP. [PR/277310: This issue has been resolved.]
• In JUNOS Release 9.0 and later, the multicast snooping process (snoopd) leaks memory even if it is not configured. As a workaround, disable the process if it is not required. [PR/279378: This issue has been resolved.]
• Under certain circumstances, an empty pair of parentheses "( )" appears after the value in the "AS path:" field of the output from the show route extensive command. [PR/281023: This issue has been resolved.]
• When both of the following conditions apply, BGP evaluation of alternate multipaths does not work correctly: (1) an IBGP peer and an EBGP peer both provide the same prefix with the same AS path, and (2) the configuration for the EBGP peer includes the multipath multiple-as statement at the [edit protocols bgp group group-name] hierarchy level but the IBGP peer's configuration does not. [PR/281447: This issue has been resolved.]
• When two Protocol Independent Multicast (PIM) any-source multicast (ASM) routers on a LAN segment have a directly attached receiver to this segment, IIF_MISMATCH error messages might be displayed preventing the creation of an (S,G) state. This problem occurs with PIM and Multicast VPN configurations. [PR/281662: This issue has been resolved.]
• Using PIM, certain multicast routing topologies might cause delays in multicast route convergence. [PR/282109: This issue has been resolved.]
• Sometimes memory blocks can become corrupted due to an invalid write into free memory. [PR/283819: This issue has been resolved.]
• A bidirectional forwarding detection (BFD) protocol session might become stuck for different reasons depending on the release in question. For JUNOS Release 8.3 or earlier, the BFD session might be in the failing state. For JUNOS Release 8.3 or later the BFD session might be in init state. As a workaround, issue the clear bfd session command to bring the session back up. [PR/286331: This issue has been resolved.]
• When BGP multipath is enabled (the multipath statement is included at the [edit protocols bgp group group-name] hierarchy level) and route updates arrive from multipath and nonmultipath peers in a certain order, load balancing across paths might stop working correctly. [PR/288694: This issue has been resolved.]
• When BGP deletes a secondary route, the routing process (rpd) might exit unexpectedly and dump core. [PR/290863: This issue has been resolved.]

MPLS Applications

• On M-series and T-series routing platforms, if MPLS traffic is being forwarded on the secondary path of an LSP when the primary path is also functional, the "Traffic statistics" section of the output from the monitor label-switched-path lsp-name command might show incorrect values. [PR/80591: This issue has been resolved.]
• On an M120 router, the ping mpls rsvp command fails when an LSP is configured for link protection (the link-protection statement is included at the [edit protocols mpls label-switched-path lsp-name] hierarchy level) and traffic is being routed through the bypass LSP. [PR/233693: This issue has been resolved.]
• On M- or T-series routers, when an MPLS LSP gets optimized, the MPLS MIB counters associated with the path change event are not updated. [PR/265931: This issue has been resolved.]
• After upgrading to JUNOS Release 8.4 or later, LDP neighborship is not established with another vendor's equipment because of a subnet mismatch. The fix adds a new configuration statement, allow-subnet-mismatch, that ignores subnet mismatch for the source address in LDP link hello packets. [PR/285933: This issue has been resolved.]
• When the target of the ping mpls rsvp command is another vendor's router, the value in the Local transmit time field is a UNIX timestamp instead of an NTP timestamp as specified by RFC 4379. [PR/289535: This issue has been resolved.]
• Packet loss can occur following an RSVP auto-bandwidth adjustment. [PR/289553: This issue has been resolved.]
• Other vendor implementations might send status TLV notification messages with the U-bit set to 0 and F-bit set to 1. While such a combination is not recommended according to RFC 5036, the JUNOS software will tear down the LDP session upon receiving such a status TLV message. [PR/290845: This issue has been resolved.]
• The autobandwidth adjustment intervals for label-switched paths (LSPs) do not conform to the configured values. [PR/297771: This issue has been resolved.]

VPNs

• When you reboot the tunnel interface or the FPC holding the tunnel interface, the virtual tunnel interface is not recreated. As a workaround, deactivate and activate the virtual tunnel interface configuration. [PR/266170: This issue has been resolved.]
• When the tunnel-services statement is configured at the [edit routing-instance instance protocols vpls] hierarchy level and a VPLS interface is configured with an MTU, the virtual tunnel interface might flap due to unrelated configuration changes. As a workaround, remove the tunnel-service statement in the routing instance configuration. [PR/297141: This issue has been resolved.]

Class of Service

• Adding and deleting an interface many times by configuring the scheduler-map-chassis statement at the [edit class-of-service interfaces ge-1/1/*] hierarchy level might cause a memory leak in the class-of-service process (cosd). As a workaround, restart the class-of-service process (cosd). [PR/82546: This issue has been resolved.]
• When you configure an EXP classifier for a routing instance (include the exp classifier statement at the [edit routing-instances routing-instance-name classifiers] hierarchy level), the class-of-service process (cosd) might generate a core file. [PR/101490: This issue has been resolved.]
• On M120, M320, and MX-series routers, if the value set by the transmit-rate statement at the [edit class-of-service schedulers scheduler-name] hierarchy level is larger than the value set by the buffer-size statement at that level, forwarding latency is greater than expected. [PR/233213: This issue has been resolved.]
• On the MX960, bandwidth sharing across high priority and strict-high priority schedulers might not be as expected. This issue occurs when the schedulers are configured on logical interfaces. [PR/265603: This issue has been resolved.]
• On M120 routers, MX-series routers, and on M320 routers with E3-FPCs, MPLS transit traffic with a label stack that performs a pop operation at the penultimate node is not shaped according to the configured transmit rate exact value, which results in more traffic being sent than should be allowed. [PR/282002: This issue has been resolved.]

Forwarding and Sampling

• On M120 and MX-series routers, if you configure both a firewall filter and interface sampling for ingress traffic on the same interface (by including both the filter and sampling statements at the [edit interfaces interface-name unit logical-unit-number family inet] hierarchy level), the interface discards all incoming packets. As a workaround, implement input sampling as an action in the then section of a firewall filter. [PR/103206: This issue has been resolved.]
• When a firewall policer that has the loss-priority action in its then clause is applied to multiple outbound interfaces, loss priority might not be set correctly at all interfaces. [PR/274346: This issue has been resolved.]
• At the [edit routing-options flow route route-name] hierarchy level, when a range of numeric values is selected for firewall filtering using a statement, the match condition is ignored. Such statements typically include destination-port, port, protocol, and source-port. [PR/275650: This issue has been resolved.]

Routing Policy and Firewall Filters

• When a large number of firewall filter counters are configured, the Packet Forwarding Engine (pfed) process might panic and generate a core file. The pfed process is responsible for Packet Forwarding Engine statistics collection and does not affect traffic forwarding. [PR/265256: This issue has been resolved.]
• PPTP transit traffic is being dropped when the dhcp-relay statement is configured. The root cause is the implicit filter used for snooping that is not specific enough and only matches on the port number where the protocol UDP should also be matched. [PR/279432: This issue has been resolved.]
• On the MX platform, a firewall filter with the ip-options statement included and applied to the loopback interface might not operate correctly. [PR/283215: This issue has been resolved.]

Network Management

• A syntax error in the mib-rfc3811.txt MIB file prevents SNMP from using the MIB. The file is included in the package accessible at https://download.juniper.net/ software/junos-export/release/juniper-mibs-release-signed.tgz, where release is a JUNOS Release number such as "8.2R4.5". [PR/80648: This issue has been resolved.]

Release 9.0R2

The following issues have been resolved since JUNOS Release 9.0R2. The identifier following the description is the tracking number in our bug database.

Software Installation and Upgrade

• When upgrading from JUNOS Software Release 9.0R1 to JUNOS 9.1B1, the new master Routing Engine is upgraded; however, the old master (new backup) Routing Engine is not upgraded. As a workaround:
  • Deactivate the redundancy graceful-switchover statement at the [edit chassis] hierarchy level.
  • Deactivate the nonstop-routing statement at the [edit routing-options] hierarchy level.
  • Use the request system add command to upgrade the old master Routing Engine to JUNOS Release 9.1B1. [PR/272781: This issue has been resolved.]

Platform and Infrastructure

• Packet Forwarding Engine (PFE) notification statistics for the "options or ttl expired (not RE-destined)" counter are increased incorrectly. [PR/64951: This issue has been resolved.]
• During a Routing Engine switchover, the Flexible PIC Concentrator (FPC) might reset multiple times. [PR/70857: This issue has been resolved.]
• Due to changes in the JUNOS TCP/IP networking stack, the output of the show connections command may be different from JUNOS Release 8.4 and earlier. [PR/103330: This issue has been resolved.]
• The PCAP support for Multilink Point-to-Point Protocol (MLPP), Multilink Frame Relay (MLFR), and Multilink Frame Relay (MFR) is not available. [PR/239642: This issue has been resolved.]
• On some Routing Engines, the smartd process may display the following error: “atastandbyarmset.” [PR/253775: This issue has been resolved.]
• Using the ox option with the smartd process is not recommended on mounted devices because it may result in unexpected behavior. [PR/255473: This issue has been resolved.]
• If there are a lot of aggregate next hops and BGP routes pointing at some of them, a quick link flap combined with the BGP route churn might cause the Packet Forwarding Engine to restart unexpectedly. [PR/268204: This issue has been resolved.]
• With certain traffic patterns, MX-series and M320 routers with 3.0 forwarding ASICs might experience packet loss. To recover, you must reboot the affected Dense Port Concentrator (DPC). [PR/268274: This issue has been resolved.]
• On a J-series router configured with an IPSec tunnel, Ethernet frames smaller than 64 bytes might not pass through. [PR/268965: This issue has been resolved.]
• In JUNOS software Release 8.5R1 only, when the router receives an MPLS LSP ping packet, a kernel memory leak is triggered in the network packet buffer and, upon exhaustion, packet transfer is not possible between the Routing Engine and the Packet Forwarding Engine. [PR/273024: This issue has been resolved.]
• On a T1600 router, the Link Aggregation Control Protocol (LACP) does not work on the following Ethernet interfaces: 10x1 Gigabit Ethernet, IQ2, or Fast Ethernet. [PR/274586: This issue has been resolved.]

User Interface and Configuration

• If you use the NETCONF API to modify the configuration database when it has been locked by another NETCONF session, or if you try to delete a configuration statement that does not exist, the NETCONF server returns both the <rpc-error> and <ok/> tag elements as children of the <rpc-reply> tag element. [PR/62664: This issue has been resolved.]
• JUNOS devices cannot be managed by Session and Resource Control (SRC) software. [PR/273117: This issue has been resolved.]
• When you use the configure private command and then change two static routes at the same time at the [routing-options static] hierarchy level, the commit may fail. As a workaround, configure one static route at a time. [PR/273251: This issue has been resolved.]
• When the MX240 DC chassis is in a slightly elevated temperature environment with the fans in high speed, the show chassis environment command might hang. As a workaround, terminate the CLI session by typing CTRL+C to open a new CLI session. [PR/274015: This issue has been resolved.]

Interfaces and Chassis

• On M20 routers, when you start the router with Routing Engine 0 and System and Switch Board (SSB) 0 as master components, issue the request chassis routing-engine master switch command, and then log in to Routing Engine 1 and issue the request chassis ssb master switch and request system reboot commands, the ONLINE LED might remain lit on both SSBs. [PR/74283: This issue has been resolved.]
• When you configure the default-address-selection statement at the [edit system] hierarchy level, Routing Engine graceful restart may cause gateway GPRS support node (GGSN) services to be unreachable. [PR/232197: This issue has been resolved.]
• Input statistics for aggregated Ethernet interfaces incorrectly report zero, regardless of input traffic volume. [PR/266271: This issue has been resolved.]
• The chassis LED status returned by the MIB jnxLEDState does not reflect the actual chassis alarm LED. [PR/266326: This issue has been resolved.]
• SFPCs caused the warning message “WARNING: Unknown FPC 0x1f4” when checking PIC compatibility. [PR/266854: This issue has been resolved.]
• When there is an Address Resolution Protocol (ARP) entry for Virtual IP (VIP), Virtual Router Redundancy Protocol (VRRP) might not respond to ARP requests for VIP while transitioning to master state. [PR/268627: This issue has been resolved.]
• An error in the chassis process (chassisd) causes a small memory leak when the show chassis hardware extensive command is executed. [PR/268925: This issue has been resolved.]
• On an MX platform, when a lot of traffic is going to the Routing Engine and the route changes, the Dense Port Concentrator (DPC) might trigger an assertion without producing a core dump. [PR/269699: This issue has been resolved.]
• When you configure multiple interfaces with the vlan-id-range statement and cover a large number of VLAN IDs, the Dense Port Concentrator (DPC) might restart unexpectedly. [PR/271456: This issue has been resolved.]
• In Release 9.0R1, you can configure only 4096 Layer 2 Tunneling Protocol (L2TP) sessions with shapers and policers. [PR/273804: This issue has been resolved.]

Routing Protocols

• Kernel replication might fail when a pending change on a virtual loopback tunnel (VT) interface is replicated. [PR/69862: This issue has been resolved.]
• The redistribution of OSPF point-to-point (P2P) LAN interfaces from a routing instance into the main routing instance can fail, displaying the following log message: “Jan 23 15:24:50 router rpd[6063]: cannot perform nh operation ADDANDGET nhop 0.0.0.0 type unicast index 0 errno 45.” As a workaround, redistribute the interfaces routes within the routing instance into the main routing instance. [PR/271130: This issue has been resolved.]
• The routing process can restart unexpectedly if it receives a BGP flow NRLI specification with an undefined subcomponent type. [PR/274421: This issue has been resolved.]

MPLS Applications

• The system can leak next-hop resources when RSVP link protection is enabled. [PR/265295: This issue has been resolved.]

VPNs

• An interoperability issue exists between NetScreen-specific features (NHTB) and JUNOS Enhanced Services. [PR/274937: This issue has been resolved.]

Class of Service

• On MX-series Ethernet Services routers, if you apply class of service (CoS) to an integrated routing and bridging (IRB) interface within a VPLS domain using virtual loopback tunnel (VT) interfaces, traffic from remote customer edge (CE) routers might not be forwarded during a graceful Routing Engine switchover. As a workaround, modify your VPLS domain to use LSI interfaces by including the no-tunnel-services statement at the [edit routing-instances instance-name protocols vpls] hierarchy level. [PR/252468: This issue has been resolved.]
• On MX960 routers, the class-of-service process does not provide information about SNMP objects whose names begin with "jnxCosQstat." As a result, SNMP queries on those objects fail with an error message. [PR/269419: This issue has been resolved.]

Forwarding and Sampling

• If a term in a firewall filter specifies a range of values for a source or destination address or port, the filter might not match packets as expected. As a workaround, define the addresses and ports explicitly. [PR/265023: This issue has been resolved.]

Release 8.5R2

The following issues have been resolved since JUNOS Release 8.5R2. The identifier following the description is the tracking number in our bug database.

Platform and Infrastructure

• The MultiServices PIC might not work correctly when the PIC is loaded and frequent commands related to the PIC are issued. [PR/81826: This issue has been resolved.]
• When IPSec is configured on a logical interface and the protocol family is IPv6, graceful Routing Engine switchover (GRES) might fail if an MTU change is attempted on that interface. [PR/230128: This issue has been resolved.]
• J-series multilink interfaces behave well when fragments are in round-robin fashion and arrive in order. However, if fragments are out of order, then they will suffer some latency and packet loss during reassembly. [PR/240019: This issue has been resolved.]
• When using file copy FTP, the IP address specified as the source address is not used for establishing a connection with the peer FTP server. [PR/240580: This issue has been resolved.]
• When graceful switchover and RLSQ interfaces are configured, the kernel generates a core file on the backup Routing Engine with the error message "panicstr: rnh_index_alloc: nhindex 116435 could not be allocated." [PR/241502: This issue has been resolved.]
• When an unnumbered Ethernet interface has a loopback address as a donor and if the address configured on the loopback is a subnet address, a ping to the subnet address does not work. [PR/253804: This issue has been resolved.]
• In some situations, the interface counter account doubles the number of packets. [PR/253946: This issue has been resolved.]
• The clear arp command does not function for logical routers. [PR/253957: This issue has been resolved.]
• The show arp command does not function for logical routers. [PR/253958: This issue has been resolved.]
• A router running multicast over aggregate SONET or aggregate Ethernet interfaces could experience a Packet Forwarding Engine crash when a constituent link flaps. [PR/257691: This issue has been resolved.]

User Interface and Configuration

• If you use telnet to connect to the JUNOScript Perl module, the connection fails if the password or login name includes special characters. [PR/241236: This issue has been resolved.]
• Certain JUNOScope wizards (devices, groups, users, schedules, and RADIUS configuration ) do not work with Netscape 7.0. The workaround is to use Netscape 6.2. [PR/260326: This issue has been resolved.]

Interfaces and Chassis

• If a low-speed bundle is congested, the jitter for link fragmentation and interleaving (LFI) traffic is high even though fragmentation is configured for Multilink Point-to-Point Protocol traffic. [PR/77862: This issue has been resolved.]
• If you clear IPv6 statistics, deactivate IPv6 route accounting at the [edit forwarding-options family inet6 route-accounting] hierarchy level, and resume IPv6 traffic across an interface, the "Input bytes" and "Input packets" fields in the output of the show interfaces extensive command might display incorrect values. [PR/99461: This issue has been resolved.]
• On the M120 and MX-series platforms and the M320 Enhanced III FPCs only, forwarding IPv6 transit packets might stop transmitting traffic but still be able to receive traffic. All packets going out this interface will be dropped. To recover, reboot the FPC on M320 platforms, the DPC on the MX-series, or the FEB on the M120 platform. There is no workaround. [PR/105266: This issue has been resolved.]
• For Gigabit Ethernet interfaces on J-series Services Routers, the link-mode and speed statements at the [edit interfaces ge-fpc/pic/port] hierarchy level are mutually dependent; that is, if you include one, you must include the other. If you do not, the interface process generates a warning and uses autonegotiated values. For Gigabit Ethernet interfaces on other routing platform types, the speed statement is not available, so including the link-mode statement alone is valid. Nevertheless, the interface process writes the following message to its log and the system log: "Speed and linkmode duplex settings are mutually required." (Note further that the ink-mode statement is actually nonoperational on non-J-series routing platforms, because the only valid value for it is the default, full-duplex.) [PR/228857: This issue has been resolved.]
• The "accept data" warning message for VRRP might not correctly display the logical unit identifier. [PR/236135: This issue has been resolved.]
• When a PIC detachment process takes a long time for an IQ2 PIC (for example, when a large number of route updates are triggered after an IQ2 PIC is brought offline), the PIC chassis process (pic-chassisd) connection might not be closed properly. [PR/239944: This issue has been resolved.]
• On the J4350 and J6350 Services Routers, the show chassis interfaces extensive command does not display "Carrier Transitions" and certain other statistics properly. [PR/241086: This issue has been resolved.]
• Padded MPLS-encapsulated IPv4 packets that exit an LSP can cause the egress interface to stop forwarding packets. This can happen when the router is configured as a VPN PE, or when the router is the penultimate node of an LSP. The problem only occurs when the packet has been padded to meet the minimum Layer 2 frame size (for example, Ethernet media require frames to be a minimum of 64 bytes long). This issue is applies to the M120 and MX-series platforms and to the M320 Enhanced III FPCs. To recover, reboot the FPC on the M320, the DPC on the MX-series, or the FEB on M120 routing platforms. [PR/251042: This issue has been resolved.]
• The ATM PIC driver might not always use the minimum port shaping rate (of all the ports on a multiport ATM DS3 or E3 PIC) selected for cell transmission shaping, in situations where the DS3 or E3 port parameters are not identical on all ports of a multiport ATM DS3 or E3 PIC. The PIC shaping rate is always updated to conform to the last port setting updated by the PIC software driver, rather than using the minimum port (shaping) rate. There is no syslog message to inform the user of the shaping rate decision applied by the software driver. [PR/252837: This issue has been resolved.]
• On the J-series Services Routers, on 4-port Fast Ethernet Enhanced Physical Interface Modules (ePIMs), interfaces might stop working correctly and lock up a port when operating in half-duplex mode. As a workaround, hard code the link-speed and link-mode to 100m full duplex. [PR/253329: This issue has been resolved.]
• When you configure a logical interface with encapsulation ether-vpls-over-atm-llc, the packets destined for to the Routing Engine are dropped by the Packet Forwarding Engine. [PR/255713: This issue has been resolved.]
• When a Routing Engine assumes mastership, it attempts to reconnect to the Packet Forwarding Engine. The Packet Forwarding Engine sends all the information to the master Routing Engine. The new master Routing Engine then attempts to retrieve SFP information from the PIC, but the PIC fails to send it. [PR/256032: This issue has been resolved.]
• On an M40e router with 4 MS-100 PICs and 4 ChOC12s, the SONET interface might display remote defect indication (RDI) or alarm indication signal (AIS) alarms when the router is rebooted. [PR/257419: This issue has been resolved.]
• When both PIM and OSPF are configured on an IQ2 PIC, OSPF may lose adjacency if protocol PIM is removed. [PR/257848: This issue has been resolved.]
• After a graceful Routing Engine switchover on the M10i router, alarms might not resynchronize to the new primary Routing Engine. [PR/258034: This problem has been resolved.]
• On an MX-series router with a 4-port Gigabit Ethernet DPC, when you configure asynchronous notification, it does not function properly. [PR/259304: This issue has been resolved.]
• On the J2320 and J2350–series Services Routers, Gigabit Ethernet interfaces sometimes stop transmitting the packets if Transmit Descriptors are not updated with Transmit Done status properly. To work around this issue, restart pic 0. [PR/261010: This problem is resolved.]

Services Applications

• When you configure twice NAT with static source and static destination translation, the destination port for ICMP flows might change (the ports are supposed to remain unchanged). [PR/96701: This issue has been resolved.]
• If Network Address Port Translation (NAPT) is configured and multiple short-lived flows are established, ports on AS PICs might not be assigned correctly. In some cases, this situation causes the AS PIC to stop functioning. [PR/229287: This issue has been resolved.]
• On the MultiServices 400 PIC, a memory warning flag might be set even with low traffic rates. [PR/251908: This issue has been resolved.]
• The show services pgcp active-configuration command does not display byte units for the "MG maximum PDU size" and "MGC maximum PDU size" output fields. [PR/256801: This issue has been resolved.]

Routing Protocols

• Trace pointers for some BGP tasks were not updated on reconfiguration. [PR/69321: This issue has been resolved.]
• The show pim source does not display the correct information for both direct and non-direct sources. [PR/253629: This issue has been resolved.]
• PIM anycasts do not work when the source is connected to the Rendezvous Point router. [PR/256637: This issue has been resolved.]
• Route target filtering breaks when the last community received is withdrawn, because the route-filtering logic is being bypassed. Absence of any route target received from the peer is being treated as if 0/0 default was received from the peer. [PR/257011: This issue has been resolved.]
• The show multicast snooping route bridge-domain name source-prefix prefix/length command causes the multicast snooping process (mcsnoopd) to stop functioning. The workaround is not to use the source-prefix option. [PR/257788: This issue has been resolved.]
• The routing processes can generate nonfatal core dumps. [PR/258134: This issue has been resolved.]
• Deconfiguration of a routing instance on a router configured with uRPF may cause the routing process to restart. [PR/259727: This issue has been resolved.]

MPLS Applications

• A router configured with a point-to-multipoint transmit-switch connection might stop functioning if the transmit label-switched path of the connection flaps. [PR/229175: This issue has been resolved.]
• You might encounter an interoperability issue when Cisco IOS-XR or IOS includes the node-id sub-object as part of the RRO in Reservation messages. The JUNOS software is unable to find the next-next-hop router's interface address to signal a node-protecting bypass LSP. [PR/237491: This issue has been resolved.]
• MVPN P2MP deactivation of a vt interface in a particular VPN, say VPN-A on receiver PE, affects multicast traffic forwarding in other VPNs for a few seconds. This issue is also experienced during the activate sequence of the vt interface in VPN-A. [PR/252697: This issue has been resolved.]
• If an MPLS LSP configured with fast reroute is not advertised into the IGP, that LSP might reuse an old unicast list and cause traffic drops. [PR/253352: This issue has been resolved.]
• If the authentication method for a LDP session is changed from using authentication-key to using authentication-key-chain or vice-versa, other unrelated LDP sessions may flap in addition to the affected LDP session flapping. [PR/258395: This issue has been resolved.]
• If there are many LSP flaps, used for the p2mp-receive-switch, and graceful restart is enabled, the interface might stop forwarding traffic. To recover, deactivate and then activate the protocol connection p2mp-receive-switch configuration. As a workaround, disable graceful restart. [PR/264930: This issue has been resolved.]

VPNs

• When configuring multicast VPN for point to multipoint, the tunnel-limit statement for dynamic selective provider tunnels does not function. [PR/250701: This issue has been resolved.]
• A receiver directly attached over an Ethernet connection to a sender PE configured with multicast VPN over point-to-multipoint fails to receive multicast traffic. No such issue is observed when the receiver is connected over a SONET or ATM link. [PR/252314: This issue has been resolved.]
• When you configure the vlan-tags statement under routing instances for a VPLS or a virtual switch instance, the JUNOS software might produce commit check error messages. The workaround is to use the vlan-tags statement in the configuration for interfaces only. [PR/256958: This issue has been resolved.]

Class of Service

• On the M320 and T-series routing platforms, when you map multiple forwarding classes to the same queue and then include the multiple of those definitions in the scheduler map, the configuration might fail. [PR/103370: This issue has been resolved.]
• The behavior aggregate (BA) classifier is not applied to a logical interface configured with the encapsulation ether-over-atm-llc statement. [PR/255742: This issue has been resolved.]
• For IQ PICs on the M-series and T-series routers and Enhancing Queueing DPCs on the MX-series routers, when you configure a scheduler map with a queue configured with priority strict-high, in certain situations such as when a PIC is bounced, the incorrect queue buffer might be calculated. [PR/256263: This issue has been resolved.]
• When a PIC goes offline and then online, the following message might display on the Packet Forwarding Engine console: “cosman_compute_mad_state: No ifd for ifd_index <ifd index value>.” The message does not indicate any effect on the operation of the router unless a temporal or delay buffer configuration is present on the router. [PR/257814: This issue has been resolved.]
• If you configure a traffic shaper of over 75 MB on a Gigabit Ethernet interface, its overall throughput might decrease. For Gigabit Ethernet interfaces, configure shapers with a size of less than 75 MB. [PR/257951: This issue has been resolved.]

Forwarding and Sampling

• If you configure a policer with a burst size limit larger than 67 MB, interfaces to which the policer is applied might not forward traffic. On some platforms the limit is higher, and the limit also depends on the available bandwidth. [PR/99758: This issue has been resolved.]
• Firewall filter source-address match sometimes does not function properly because of corner cases during firewall optimization. The workaround is to rearrange the term order of the filter. [PR/262491: This issue has been resolved.]

Network Management

• When you configure an SNMP client list with a logical router, clients are not restricted. [PR/254574: This issue has been resolved.]

Copyright © 2008, Juniper Networks, Inc. All rights reserved. Trademark Notice.