Open Issues

On all platform with BGP segment routing-traffic engineering (SR-TE), in a segment routing topology, the transit IPv4 traffic might miss labels and might get dropped in first hop, when the ingress forwards the traffic. It might miss all the labels except the last hop in the IPv4 traffic forwarded by the next hop interface. PR1505592

The FPC core files get stuck in /var/lib/ftp/in/. Sometimes when core files from the multiple FPCs are dumped, core files from one FPC might get left behind in the /var/lib/ftp/in and are not processed by core-mgr. If the show system core-dumps shows some older core files in the /var/lib/ftp/in/fpc-hostname/ and core files in the /var/core/node-name is newer, then it indicates this issue is hit. PR1522404

GRES is not supported when the FPCs are restarted. PR1539685

The dp_xx_viqm_intr_req_enq_err error might be observed intermittently in the logs and it can be safely ignored as it is observed mostly for a number of scenarios which does not have any service impact. Because of the design of the underlying silicon, this interrupt cannot be completely avoided which does not have absolutely no service impact in almost all scenarios unless joined by a few other severe interrupts. PR1553943

When an aggregated Ethernet link is brought down, the following transient error message appears: [Error] Nexthop: EalNhHandler: failed to add Nh: xxxx, type: composite, as pil add failed. There is no functional impact due to this error. PR1570710

When the scheduler configuration is not applied to all 8 egress queues of an interface and one or more egress queues have the buffer size remainder configuration, the distribution of buffer to egress queues with buffer size remainder is not distributed correctly, which might lead to an unexpected tail drop. PR1575798

On PTX10000 line of routers, CB slot becomes Fault Standby after issuing the request node power-off re slot command on the primary re slot. The correct CB state is offline. PR1581476

If the input subscription paths contain a ":" character, the extension header in case of GNMI and certain fields for the show network-agent statistics command shows incorrect values. PR1581659

When a switchover is performed at the time when the protocol sync in progress, there is a remote possibility that you might hit the bgp_read_message:3261: NOTIFICATION sent to : code 1 (Message Header Error) subcode 2 (bad length) value 27395 error and session flaps. PR1589283

Removing or adding nested firewall filter in lo0 sets/clear displays the following error: Performing action log for error /fpc/0/evo-cda-bt/0/cm/0/btchip/0/filter_action_0_intr_pmv_eq_zero (0x4514e4) in module: btchip with scope: pfe category: functional level: minor. PR1589296

The ECMP tracer fails to count output packet on the aggregated Ethernet interface. PR1597038

Brief multicast traffic loss is observed while performing Routing Engine switchover with GRES. PR1593810

After configuring the warm-standby option, wait for 3 minutes before performing a Routing Engine switchover. PR1623601

When all Packet Forwarding Engines go offline and are brought online with multicast route being active during this cycle, multicast traffic is lost permanently due to the absence of mcast route. PR1598894

On PTX10008 platforms, ZTP fails on 40G WAN ports intermittently. PR1602131

The sFlow ingress sampling reports incorrect OIF and next hop with user IPv6 traffic in an ECMP scenario at last hop router with the ultimate hop pop LSP.PR1602448

On the PTX10008 platform, fan tray controller removal or absence alarm is not generated. PR1605987

If rpd agent sends indirect next hop deletions or additions in out of order to backup rpd, the rpd generates core file. This is a backup rpd crash issue and does not impact any functionality. PR1607553

In certain cases when an interface is deleted and the corresponding next hop is about to be deleted. At this point, rpd restarts or GRES happens, then the following error message is displayed: RPD_KRT_KERNEL_BAD_ROUTE: krt unsolic client.: lost ifl 0 for route. This does not have any functional impact. The error indicates that the rpd is replying the route whose next hop's interface is now delinked and linked to a local interface. PR1612487

After system reboot or power cycle, a few fabric links showing init CRCs is expected with Gen 3 and Gen 5 fabric connectors. They do not indicate any issue as long as the CRC count is not incrementing with time. A higher corrected Tmax value after system reboot is not an indication of bad link. If you see links in FAULT or links showing incrementing CRCs after system reboot, do a graceful OIR of the affected FPC. PR1613214

With Gen 3 fabric connectors, it is possible that a few fabric links come up in FAULT or show incrementing CRCs. Normally, this type of issue starts happening after a FPC/SIB reseat event. And it might continue even after FPC goes offline/online or system reboot. Do gracefully reseat of the affected FPC to recover from the issue. PR1613559

After the FPC goes offline, minor cm-errors might occur for ZFI block on other FPCs. These errors can be cleared and does not affect system functionality. PR1616179

The Routing Engine switchover silently, so the automatic switchover has been disabled by default. To enable it, use the set chassis redundancy failover command and to disable it again, use the delete chassis redundancy failover. If a user migrating from Junos OS to Junos OS Evolved, then it is possible that they will have the chassis redundancy failover configuration which is not operational in Junos OS Evolved, however we need to explicitly delete redundancy failover using the CLI. PR1617720

MPLS family filters do not work when it is applied to passive monitoring interfaces. PR1620470

Incorrect sensor modeling/mapping when using /junos/system/linecard/interface/ native telemetry streaming. PR1621037

On a scaling system, load override and commit of baseline cfg cause the rpd to spin high on CPU. As GRPC configuration is removed, sensors need to be uninstalled. But, the rpd does not respond to these telemetry sensors uninstallation requests, so sensors uninstallation fails. Later, when GRPC is enabled back on box and the same sensor profile (cfg.json file used with jtimon) is requested from collector, the rpd sends packets with higher sequence numbers (because sensor was not removed from rpd earlier) and this is considered as drops by collectors which rely on sequence-numbers. PR1621347

The [Error] Nexthop: Egress NhChain: numOfTags is 2 and srteGlobalIndex is 0 error is not seen until there is a composite next hop with 2 labels in it. There is no impact in behavior and flow of traffic with these errors. PR1621689

The CDA application crashes as soon as FPC booted up resulted in not publishing pfeE object. Hence fabricPfeE objects are not created or deleted. SIB endpoints cleanup are dependent on the deletion of fabricPfeE objects. Stale SIB fabric endpoints prevent fabric link bring up in further FPC reboots. PR1624765

Unable to execute /usr/sbin/picinfo. Bad file descriptor found when the clear services inline-monitoring statistics command is issued. PR1624094

When using a source IP address as the management interface with the RPF check set to strict on the interface, the response for the ICMP ping from the peer on the management interface is dropped by the Linux kernel as it expects the path to the peer to be on the WAN side. PR1498255

The GRES triggered using the request chassis routing-engine master switch command shows the following connector driver overlay message: {master} user@host> [ 1185.081257] gpio-jnx-i2cs gpio-jnx-i2cs.50: Asserting power_status irq 59 [ 1185.125182] OF: overlay: overlay_is_topmost: #9 clashes #10 @/ftc0/i2c-bus/i2cs@54/fan_hwmon [ 1185.125183] OF: overlay: overlay #9 is not topmost. PR1539232

The licenses used value for the Port Bandwidth Usage (PAYG) field under the show system license does not update when new interfaces are created. In order to correct the value, the IFMAND process needs to be restarted to invoke a refresh. PR1595179

If the marvd daemon restarted for any reason, the device becomes unreachable because of PCI uncorrectable non-fatal AER errors getting generated and the system reboots without any core file. It happens when the marred crashes. PR1600870

In a network where there are high packet drops, if the peer end (grpc collector or client) gets closed, the TCP sessions are held in an established state in the device. PR1592542

If you configure LDP auto-targeted session after establishing IGP paths, it fails to establish an expected backup MPLS paths. PR1620262

When two trap groups are configured with the same trap destination address, snmpd process sends trap using only the first trap-group (community). PR1623201

The default remnant hold time is 300 seconds which is not sufficient for this scale (1.7M FIB routes, 37,000 LSPs, and jFlow enabled). You need to increase the hold time to 600 seconds to avoid any protocol flap during the rpd crash or restart. Use the set routing-options forwarding-table remnant-holdtime 600 to set the hold time. PR1593445

Traffic loss is observed on BGP-LU paths after restoring primary paths in setup with IGP/LDP RLFA enabled. PR1619229

When a user tries to deactivate the MPLS related configuration, the commit fails on the backup Routing Engine. PR1519367