Open Issues

IPv6 transit traffic statics output is missing. PR1662174

With dhcp-snoop configuration, jdhcpd might crash during configuration commit on Junos OS Evolved based platforms.PR1688644

On all Junos Evolved platforms, configuring Link Layer Discovery Protocol (LLDP) with system services netconf notification enabled triggers l2cpd crash. This crash causes the CPU to spike.PR1695057

In ACX7509, time error spike is seen with PTP long. runPR1697093

We observe core with rpd with BGP flowspec if secondary-independent-resolution is configured. PR1722715

We might encounter jdhcpd core during initialization. This is rare, and there is no service impact because of this core (as the process recovers immediately). PR1730717

ACX7348: 400G-AOC module support is not available in Junos OS Evolved Release 23.4. PR1732958

On Junos OS Evolved ACX platforms with GRES (Graceful Routing Engine switchover), after performing GRES switchover jdhcpd does not start on the new primary Routing Engine due to which DHCPv4/v6 (Dynamic Host Configuration Protocol) session binding is lost resulting in traffic loss. PR1740530

Servo sends some of DEL-REQ pakcets to 0:00:00:00:00:00. This is seen when moving the port from 400g to 25g/40g/100g.PR1740612

show pfe statistics traffic command displays 0 in Input packets: and Output packets: after Packet Forwarding Engine app (evo-pfemand) restart for some time. Correct values are displayed after the delay. PR1745512

During a continuous network churn, following errors related to pfe-spec filters are seen for Packet Forwarding Engine applicaiton (evo-pfemand). [Error] compName = BrcmPlusMcast", tpName = "Irb", msg = " fn = installUnknownMcastLexemEntry, failed filter install for unit = 0, bdIndex = 3.PR1748663

In ACX7509, syncE clock failure is seen on 4x100G channelized interface. PR1756587

Test chassis fpc slot <> power on or off command does not work on ACX7348 and ACX7332.PR1757288

In scaled aggregated Ethernet cases during Routing Engine switchover when backup Routing Engine becomes primary, it takes upto 20 minutes for aggregated Ethernet interfaces statistics to be correctly reflected. The aggregated Ethernet traffic flows correctly and does not see any impact.PR1757394

For heavily scaled VPLS scenarios (8K instances) , sometimes after a restart fpc command, a few MAC entries are not programmed in hardware. This issue is not seen in normal use case and with reduced scale.PR1759887

On ACX7348 and ACX7332 Series running Junos OS Evolved multiple Routing Engine switchover can cause IDEEPROM failure on FPC and PSM, which results in PSM and FPC shown as unsupported. PR1760978

In Junos OS Evolved 23.4 on ACX7348 or ACX7332 syslog messages related to PEX configuration are seen at bootup with sib-slot which should be corrected as cb-slot in later releases. Example: Current log syslog(LOG_INFO, "%s: sib%d: pexsw bringup successful", __func__, board_slot); Expected log: syslog(LOG_INFO, "%s: cb%d: pexsw bringup successful", __func__, board_slot);PR1761710

On Junos OS Evolved ACX7024 and ACX7024X platforms, any MPLS encapsulated packet with a size less than or equal to 35 bytes on the 25G/10G/1G ports [Port No 4 to 27] gets dropped. The issue can also be observed in Point-to-Point Protocol over Ethernet (PPPoE) over Layer 2circuit scenario. PR1766889

IPTV traffic is not forwarded to test agent. This is because route update comes before the filter add. Route update checks if netrounds filter is present. If so, punts the packet to netrounds queue. Otherwise it does not punt. Since filter add comes 10 minutes later, the traffic is not punted to netRounds Queue.PR1771527

RC errors can be seen with SFP-T (Copper) optics due to vendor PHY limitation. PR1771671

The following error message is seen with the delete vlan event trigger when configured with multiple VLANs with IRB and IGMP snooping/MLD snooping enabled. No functionality impact. Error] compName = "BrcmPlusMcast", tpName = "Irb", msg = " fn = removeMemberFromL3FloodMcastGrp.PR1771915

On ACX7348 and ACX7332, during debugging of LMK alarm issue. There is a gap seen in the initial sequence of HW chips. There is no functional impact due to the issue. This is an internal fix to update the initial sequence as per discussion with HW team. PR1773212

On ACX7509 platform having routing protocol with single-hop BFD (Bidirectional Forwarding Detection) configured over AE (Aggregated Ethernet) link and OAM (Operation, Administration, and Maintenance) CFM (Connectivity-Fault-Management) configured under unit 1 for the member interfaces, also if the member interfaces of the AE are in multiple line cards (one interface in 1-3 slots and other interfaces in 4-6 slots) might result in BFD session flaps and traffic impact due to programming issues. PR1776647

3% traffic drop is seen for some l2vpn instances after flapping the aggregated Ethernet interface which can be recovered by deactivating and activating that l2vpn instance in config. PR1777608

RPD core can be see when running telemetry for protocols from top of tree and doing routing instance add delete operation. PR1778103

We need to wait for the FPC and PIC to come online before the restart FPC command is executed. When the first restart command of the FPC is executed, it puts the FPC and the PIC offline and then online. Till both the FPC and PIC status is not shown online in show chassis fpc pic-status the next restart command for same FPC should not be attempted. PR1778163

MPLS FRR with VPLS after multiple switchovers causes traffic to drop on a few VPLS instances. Deactivate/activate of the affected routing instances could recover the traffic. PR1779466

On all Junos OS Evolved platforms, with Next-Generation Multicast Virtual Private Network (NG-MVPN) configured in the core and Ethernet Virtual Private Network (EVPN) in the access, multicast traffic from the EVPN sender is not forwarded to the NG-MVPN core as multicast next hop (NH) in Packet Forwarding Engine has some programming issues resulting in traffic loss. PR1781735

During multiple Routing Engine switch-overs and router reboots, management port of one of the Routing Engines came up with 100 MBPS speed instead of 1 GBPS speed. There is no functional or stability impact due to this issue. Router monitoring might face a mild performance issue due to this lower speed of management port.PR1789156

PICD core file can be seen during FPC offline/online, HA switchover, and system restart. PR1793824

On the ACX7024 platform, the tear-down rate is low. This is due to system CPU limitations.PR1659593

On all Junos Evolved platforms, a cosmetic change is made to an existing syslog message to print more information. The syslog error message that is logged when a user with wrong auth/privacy password sends a SNMP v3 request to router has been changed to add more information. From: LIBJSNMP_NS_LOG_WARNING: WARNING: Authentication failed for <username> To: LIBJSNMP_NS_LOG_WARNING: WARNING: Authentication failed for <username> , SNMPv3 query from <NMS-ip> to <Router-IP>PR1734549

BGP passing nh path's weight as 0 first and then passing nh path's weight as 1 later causes KRT to create 2 different next hops (as weight is now part of nh key) and results in installing same route in FIB as primary and backup. PR1745661

When there are multiple events, the TI-LFA library seems to go into unnecessary multiple triggers. This can sometimes lead to pre-mature transition from MLA routes to global-convergence routes. However, we don't see this issue even when we do multiple link events manually. So, this might be some thing to do with so many multiple events intentionally/unintentionally caused by test-script. There is no service impact from this PR as it is only about pre-mature reversion to normal primary path. PR1794381