Alert Type

SRN - Software Release Notification
Low/NotificationSoftware Release Notification
Low/NotificationSoftware Release Notification

Product Affected

ACX EX MX NFX PTX QFX SRX vSRX

Alert Description

Junos Software Service Release version 25.2R2-S1 is now available for download from the Junos software download site

Download Junos Software Service Release:

  1. Go to Junos Platforms - Download Software page
  2. Input your product in the "Find a Product" search box
  3. From the Type/OS drop-down menu, select Junos SR
  4. From the Version drop-down menu, select your version
  5. Click the Software tab
  6. Select the Install Package as need and follow the prompts

NOTE: Starting on August 30th, 2024, we include PR's severity with each entry. See KB86335 [juniper.net] for the definition of PR's Severity

 

Junos Selective Update (JSU) feasible

Not applicable

Call to Action

For your review

Solution

Junos Software service Release version 25.2R2-S1 is now available.

25.2R2-S1 - List of Fixed issues

PR NumberSynopsisCategory: SRX Fleming L2NG platform support
1911533
Minor		SRX1600 switching mode : table full error message displayed even the L2 table is not full
Product-Group=junos
Severity=Minor		On SRX1600 switching mode, table full error message displayed even the L2 table is not full
PR NumberSynopsisCategory: L2NG bug tracking
1919768
Minor		"SCHED: Thread 19 (BCM AGING periodic) ran for 1016 ms without yielding" message is frequently displayed
Product-Group=junos
Severity=Minor		On SRX1500 platform in L2 mode, "SCHED: Thread 19 (BCM AGING periodic) ran for 1016 ms without yielding" message is frequently displayed
PR NumberSynopsisCategory: "agentd" software daemon
1915308
Major		The nginx sessions fail when too many RPC requests are sent after reboot
Product-Group=junos
Severity=Major		On all Junos and Junos OS Evolved platforms, after a system reboot, sending a very large number of RPC (Remote Procedure Call) calls per second (20K or more) will prevent gNMI (gRPC Network Management Interface), gNOI (gRPC Network Operations Interface) and P4RT connections from being established. This will result in nginx session failures. There will be no traffic impact due to this issue.
1919448
Major		Heavy traffic flow causing SWAN agent to disconnect from the clients
Product-Group=junos
Severity=Major		In all Junos and EVO platforms, a gRPC connection between the SWAN (Software Wide Area Network) agent and the device is being closed unexpectedly. While SWAN client/JET(Juniper Extension Toolkit) client is connecting and sending data close to 3K(3000) routes at 8KB Per RPC (Remote Procedure Call) over the connection, they can hit rare race condition which can cause the clients to disconnect.
1929823
Major		The sysd process crashes with error logs upon exceeding its memory limit
Product-Group=junos
Severity=Major		On Junos OS Evolved platforms with telemetry streaming configured, the sysd (System Daemon) process crashes on the backup RE (Routing Engine), and error logs are observed when sysd memory usage continues to grow over time, and the memory limit is exceeded. This issue is observed when the 64-bit sysd consumes more memory than the configured system limit when the device has been running for a long time (typically more than 100 days) in the dual RE scenario.
PR NumberSynopsisCategory: BBE Autoconfigured DVLAN related issues
1936257
Major		PPPoE session scale from 131K to 384K in a SVLAN subscriber model
Product-Group=junos
Severity=Major		On certain MX platform, there is an issue in scaling PPPoE sessions from approximately 131K to 384K when operating in an SVLAN subscriber model.
PR NumberSynopsisCategory: BBE routing
1922536
Major		Forwarding issues for an access DHCPv6-PD or access-internal DHCPv6-IA route or both may be seen on LNS due to an incorrect route programming of such route on PFE
Product-Group=junos
Severity=Major		Drop of traffic to subscriber DHCPv6 prefixes may be observed on LNS (L2TP network server) if CPE uses IPv6 address obtained via NDRA process as the source address for DHCPv6 negotiation instead of link-local address.
PR NumberSynopsisCategory: Border Gateway Protocol
1935730
Major		BGP session teardown due to incorrect 'prefix-limit maximum exceeded' error when RIB sharding is enabled
Product-Group=junos
Severity=Major		On all Junos and Junos OS Evolved platforms, when prefix-limit or accepted-prefix-limit is configured under family route-target with rib-sharding, prefix-limit will be incorrectly reported as 'prefix-limit maximum is exceeded'. This causes the BGP session to go down immediately. This occurs because the system incorrectly counts each received route multiple times (once for the main thread and once for each shard)
PR NumberSynopsisCategory: Track PRs in BGP Flow Spec area & is part of BGP inside RPD.
1945627
Minor		BGP IPv6 Flow Specification Session Flapping When inet6-flow Is Enabled
Product-Group=junos
Severity=Minor		In certain Junos OS releases, BGP sessions may repeatedly reset when IPv6 Flow Specification (inet6-flow) is enabled. This behavior occurs when the router receives specific IPv6 Flow Specification updates that include a default (wildcard) match. Earlier releases do not exhibit this behavior, leading to a difference in observed stability when upgrading.
PR NumberSynopsisCategory: BGP Openconfig and Sensor
1934534
Critical		On-Change telemetry mode missing notifications for BGP Ext-Comm updates
Product-Group=junos
Severity=Critical		On Junos OS and Junos Evolved OS platforms, with on_change streaming telemetry subscription mode enabled, if a change is made on a BGP (Border Gateway Protocol) external community (ext-comm), the telemetry on_change module does not generate an update.
PR NumberSynopsisCategory: MX304 PSM issuues
1923135
Major		On MX304 platform repetitive logic fault alarm on both PEMs observed
Product-Group=junosvae
Severity=Major		On MX304 platforms, PEM Logic Fault alarms may intermittently appear due to I2C transaction timing limitations. PEMs require a delay between backtoback transactions; without this delay, alarms can be raised randomly on either PEM.
PR NumberSynopsisCategory: MX304 Routing Engine issues
1913540
Minor		[MX304] Certain 'cat' commands within the 'show vmhost support-info' command in RSI try to access files that are not present on the MX304, causing error messages to appear.
Product-Group=junos
Severity=Minor		As part of the RSI process, the command "show vmhost support-info" is executed, which comprehensively collects vmhost logs using various cat commands. Some of these commands attempt to access files that do not exist on the MX304, leading to the display of error messages.
1933347
Major		Continuous I/O ata timeout errors and alarms are seen for secondary SSD read accesses
Product-Group=junos
Severity=Major		On all Junos platforms with SSD (Solid-State Drive) of Innodisk and model no. is DGM28-B56M71EC1QF-B5531, alarms are observed for secondary SSD along with continuous input/output ata timeout errors for secondary disk read accesses when the SSD firmware is unable to can't handle the continuous metadata reads. No service impact will be seen as the alarms are seen for secondary SSD alone.
PR NumberSynopsisCategory: Express BX PFE L3 Features
1917717
Major		IPv4 Family Statistics for subinterfaces streaming as zero
Product-Group=junos
Severity=Major		When the CLI setting "set forwarding-options family inet route-accounting" is enabled to activate IPv4 route accounting statistics, the streaming for /interfaces/interface/subinterfaces/subinterface incorrectly reports all IPv4 counters as zero. This occurs even though IPv4 traffic is successfully forwarded and accounted for in the Packet Forwarding Engine (PFE) .
PR NumberSynopsisCategory: Class of Service
1928420
Minor		Incorrect COS scheduler-map gets attached to an AE IFL after GRES or 'cosd' process restart
Product-Group=junos
Severity=Minor		On all Junos MX platforms with line cards (MPC1-11, MS-MPC, LC2101, LC2103, LC480, LC4800, LC4802, LC9600, JNP304-LMIC), when a scheduler-map is attached to an AE (Aggregated Ethernet) IFL (Logical Interface) directly or by 'traffic-control-profile' and when this AE is in 'replicate' mode, upon GRES (Graceful Routing Engine Switchover) or on 'cosd' process restart operation, configured scheduler-map gets removed from this AE IFL and a default (Incorrect) scheduler-map gets attached to it. This causes change in the traffic flow pattern via the COS (Class of Service) queues of the AE IFL and can impact service.
PR NumberSynopsisCategory: QFX Access Control related
1934680
Major		The radius accounting interim updates is missing framed-ip-address when manually triggered
Product-Group=junos
Severity=Major		On platforms supporting dot1x authentication, when a Radius accounting Interim-Update is manually triggered from the switch (using a COA), the resulting accounting packet does not include the Framed-IP-Address attribute. However, when the same session generates an interim update via the periodic timer (10-minute interval), the Framed-IP-Address is included as expected.
PR NumberSynopsisCategory: QFX Control Plane VXLAN
1921796
Minor		Traffic dropped after priority change in VRRP with EVPN-VxLAN scenario
Product-Group=junos
Severity=Minor		On all Junos and Junos Evolved platforms, in Virtual Router Redundancy Protocol (VRRP) with Ethernet Virtual Private Network - Virtual Extensible Local Area Network (EVPN-VXLAN) scenario, when VRRP priority is changed, the VRRP virtual Media Access Control (MAC) address can remain pinned as a static entry on the remote device. As a result, MAC movement does not occur correctly, and VRRP packets are dropped on the leaf device. This impacts VRRP operation after priority changes.
PR NumberSynopsisCategory: OpenSSH and related subsystems
1922002
Major		Major alarms showing SPMB1 not online on MX2008 will be seen on 24.2R1 and later releases
Product-Group=junos
Severity=Major		On the Junos MX2008 VMhost platform with dual Routing Engines (REs), the backup SPMB {Switch Processor Mezzanine Board} (spmb1) fails to come online following a graceful switchover on releases starting when upgrading to releases starting from Junos 24.1. As a result, if the backup SPMB cannot boot, the traffic will be impacted during switchovers.
PR NumberSynopsisCategory: DNS software support.
1851909
Critical		Junos OS: SRX Series: If a specific request is processed by the DNS subsystem flowd will crash (CVE-2026-21920)
Product-Group=junos
Severity=Critical		An Unchecked Return Value vulnerability in the DNS module of Juniper Networks Junos OS on SRX Series allows an unauthenticated, network-based attacker to cause a Denial-of-Service (DoS). Please refer to https://supportportal.juniper.net/JSA106020 [juniper.net] for more information.
PR NumberSynopsisCategory: DNX VPLS
1911208
Major		VPLS drop seen post Clear OSPF Neighbors .
Product-Group=junos
Severity=Major		Traffic drop seen in VPLS stream post "clear ospf neighbor" command.
PR NumberSynopsisCategory: AAA, auditd issues
1909811
Minor		Management access via telnet disconnects under certain conditions
Product-Group=junos
Severity=Minor		On All Junos Evolved platform, management access via telnet disconnects under certain conditions.
PR NumberSynopsisCategory: eventd, syslog infra issues
1919638
Major		Certain Junos EVO applications may become unresponsive during trace file rotation
Product-Group=junos
Severity=Major		On Junos EVO platforms, frequent trace file rotation may cause certain EVO applications to become unresponsive.
PR NumberSynopsisCategory: EVPN Layer-2 Forwarding
1926818
Major		ARP resolution failure for /32 static host routes via IRB in EVPN virtual-switch routing instances
Product-Group=junos
Severity=Major		On all Junos and Junos Evolved platforms that support EVPN (Ethernet VPN) virtual-switch routing instances, ARP (Address Resolution Protocol) resolution fails for static host routes configured with a /32 mask when the next-hop interface is an IRB (Integrated Routing and Bridging) interface and the destination host resides in a different subnet. As a result, ARP entries are not installed, MAC (Media Access Control) addresses of destination hosts are not learned, and traffic destined to those hosts becomes unreachable, causing service impact.
1934165
Minor		Packets getting forwarded without control-word flag in EVPN-MPLS set-up
Product-Group=junos
Severity=Minor		On all Junos based platforms, in an Ethernet Virtual Private Network- Multi Protocol Label Switch (EVPN-MPLS) setup, when control-word configuration is present in the routing-instance, packets are forwarded on the Integrated Routing and Bridging (IRB) without the control-word flag. This causes the packets to drop at the remote end. There is no impact on data traffic.
PR NumberSynopsisCategory: EX interfaces issues
1900610
Minor		The SNMP OID for CRC errors does not increment properly
Product-Group=junos
Severity=Minor		On all Junos and Junos OS Evolved platforms, the CRC (Cyclic Redundancy Check) error statistics for interfaces are not getting updated in the SNMP console. However, these errors are correctly displayed under the port statistics.
1904884
Major		VCP link flap due to SYSPLD read failures
Product-Group=junos
Severity=Major		On EX4100 platform VCP link flap due to SYSPLD read failures and mark SFP as unplugged
1907369
Major		10-gigabit Ethernet ports stop processing traffic on EX platforms
Product-Group=junos
Severity=Major		On Junos EX4400-24MP, EX4400-48MP and EX4400-48MXP platforms, when a 10-Gigabit Ethernet Multi-Gigabit Ethernet (MGE) port (ports 36 to 47 in EX4400-48MP and EX4400-48MXP platforms, and ports 0 to 23 in EX4400-24MP platform) is connected to a Dell server, the port stops processing traffic unexpectedly. The port remains up at the switch side, but will be reported as down at the server side. Even though the cable is disconnected, the port will continue to report a link-up state.
1915222
Major		Interfaces down after Virtual Chassis reboot
Product-Group=junos
Severity=Major		On all Junos EX Series platforms, following a Virtual Chassis (VC) reboot, random interface links GE(Gigabit) and XE(10 Gigabit) intermittently appear in a down state at the Interface Descriptor (IFD) level, even though the physical link remains operational.
1932731
Major		Crash is observed during shutdown path of an EX platform
Product-Group=junos
Severity=Major		On all EX platforms during the shutdown regular sequence a crash could occur in rare scenarios generating a core log file.
PR NumberSynopsisCategory: EX4400 PFE software
1908971
Major		DHCP Snooping drops due to misrouted server packets causing intermittent issues.
Product-Group=junos
Severity=Major		Intermittent performance issues observed across sites due to anomalous DHCP behavior, where DHCP server packets are incorrectly received on downstream switch interfaces, triggering snooping drops.
PR NumberSynopsisCategory: EX4400 platform
1942085
Major		EX4400: Kernel panic and unexpected reboot on certain EX4400 platforms due to CPU erratum
Product-Group=junos
Severity=Major		On EX4400 product family platforms, device experienced an unexpected reboot and recovers automatically. The system logs indicate a kernel panic, and a VMcore file is generated. Traffic impact will be seen until the recovery. Bios version CDEN_P_EX1_00.22.01.00 fixes this issue.
PR NumberSynopsisCategory: EX POE
1930080
Minor		PoE outage observed on EX4400 in a rare scenario
Product-Group=junos
Severity=Minor		On EX4400 platforms, loss of PoE (Power over Ethernet) power is seen on ports after port bounce or even during normal operation without specific external trigger . This causes outage on all PoE ports and devices connected to PoE port will not receive power causing service impact.
PR NumberSynopsisCategory: EX Entry Level Access VC platform
1919727
Minor		The dcpfe process crashes repeatedly with continuous error logs 'smb_transfer: SMBus ioctl failed'
Product-Group=junos
Severity=Minor		On EX4400 platforms, due to an internal communication issue within the CPU controller, the system triggers repeated fxpc panics which in turn lead to the dcpfe process (Dataplane Packet Forwarding Engine) to crash and restart continuously, resulting in service disruption.
PR NumberSynopsisCategory: FIPS related issues
1905490
Major		On MX10003 with FIPS enabled, KATs failure in SMIC_QSFP28_MACSEC_TIC causes system halt
Product-Group=junos
Severity=Major		In Junos, MX10003 platforms with Federal Information Processing Standards (FIPS) enabled experience repeated halts due to Known Answer Test (KATs) failures in the SMIC_QSFP28_MACSEC_TIC crypto path.
PR NumberSynopsisCategory: Enhanced Broadband Edge support for firewall
1928462
Major		FPC crash occurs after configuration changes are made to a service-filter on specific MX platforms
Product-Group=junos
Severity=Major		FPC crash and aftd-trio core-dump will be observed on MX platforms supporting MPC10E, MPC11E, LC9600 line cards, and MX304 after configuration changes were made to a service-filter which was in use by BBE subscribers.
1931072
Major		The shared-bandwidth-policers knob applied to subscriber firewall-filter on AE is underlying interface for demux0 results in reduced throughput
Product-Group=junos
Severity=Major		On Junos OS and Junos OS Evolved platforms, MX Series routers using MPC10E, MPC11E, LC1201, LC1202, LC1301, and LC9600 line cards allocate shared-bandwidth-policer rates based on the number of active member links when the policer is applied to subscriber firewall-filter and an AE interface is used as the underlying interface for demux0. As the number of active links decreases, the effective available throughput is correspondingly reduced.
PR NumberSynopsisCategory: ACX7332 & ACX7348 Platform Software
1904131
Major		Traffic loss is seen after disabling and enabling PTP configuration
Product-Group=junos
Severity=Major		On ACX7509, ACX7100-32C, ACX7348, ACX7332, MX10k platforms with LC480 , LC4802, MX304, QFX5700, traffic loss is seen on all ports after disabling and enabling PTP (Precision Time Protocol) configuration. The services will self-recover when traffic recovers in few milliseconds.
PR NumberSynopsisCategory: Kernel software for AE/AS/Container
1938565
Major		LACP remains in collecting and distributing state after GRES on PE and interface flap with lacp-oos-on-ndf enabled
Product-Group=junos
Severity=Major		On Junos platforms supporting Ethernet Segment Identifier (ESI) based Aggregated Ethernet (AE) Link Aggregation Group (LAG) deployments, Link Aggregation Control Protocol (LACP) does not return to the expected Attached state when a Graceful Routing Engine Switchover (GRES) occurs on the Provider Edge (PE) device and is followed by an AE member interface flap on the connected device, with lacp-oos-on-ndf (LACP out-of-sync on non-designated forwarder) enabled on the PE.
PR NumberSynopsisCategory: Integrated Routing & Bridging (IRB) module
1933452
Major		ICMP ping with higher MTU size fails in VPLS when IRB MTU exceeds MPLS core MTU
Product-Group=junos
Severity=Major		On all Junos OS platforms, configured with VPLS (Virtual Private LAN Service) and IRB (Integrated Routing and Bridging ) interfaces, when the MTU (Maximum Transmission Unit) configured on the IRB or CE interface is larger than the MPLS ( Multiprotocol Label Switching ) core interface MTU (including MPLS label overhead), ICMP Echo Request (ping) packets with higher packet sizes fail.
PR NumberSynopsisCategory: jdhcpd daemon
1927449
Major		Jdhcpd cores generated after upgrade
Product-Group=junos
Severity=Major		jdhcpd cores seen after upgrade with dhcpv6 for IANA/PD
1939685
Minor		Insert command not working for "dhcp-local-server group" hierarchy
Product-Group=junos
Severity=Minor		Insert command not working for "dhcp-local-server group" hierarchy, below error is seen: user@host# insert system services dhcp-local-server group servers-1 after syntax error, expecting `after' or `before'. user@host# insert system services dhcp-local-server group servers-1 bef syntax error, expecting `after' or `before'.
PR NumberSynopsisCategory: SRX power-mode (PMI/PME)
1858490
Major		flowd crashes due to a timing issue during IPsec SA re-keying on certain SRX platforms
Product-Group=junos
Severity=Major		On certain SRX platforms, the flowd process crash is observed during Internet Protocol Security (IPsec) Security Association (SA) re-keying. This occurs due to an internal timing issue, leading to IPsec tunnel establishment failure, traffic loss during re-key events, and traffic switchover.
PR NumberSynopsisCategory: Firewall Network Address Translation
1933239
Critical		The FPC restarts on SRX series platforms when session-persistence-scan is configured
Product-Group=junos
Severity=Critical		On Junos OS SRX Series platforms with 'session-persistence-scan' and NAT46 or NAT64 configured, modification to source Network Address Translation (NAT) rule will cause Flexible PIC Concentrators (FPCs) to restart when there is live IPv6 traffic. This will cause all traffic to be dropped and cause service disruption.
PR NumberSynopsisCategory: Firewall Policy
1932245
Minor		NSD main thread delay during policy file serialization on SRX5K platforms leads to instability
Product-Group=junos
Severity=Minor		On SRX5K platforms, during policy configuration processing using the policy file serialization feature, the Network Security Daemon (NSD) main thread can stop yielding for more than one second. This results in warning messages in system logs and can impact protocol/FPC stability if the delay becomes excessive.
1939433
Major		Flow sessions not synchronizing from primary to backup in SRX High Availability clusters
Product-Group=junos
Severity=Major		On SRX devices operating in a High Availability cluster with logical-systems enabled, the backup node may display fewer flow sessions than the primary node due to incomplete flow session synchronization. Sessions that do not synchronize to the backup node are lost during a failover, resulting in service interruption for the affected flows.
PR NumberSynopsisCategory: IPSEC/IKE VPN
1937239
Major		Srxpfe process crash during IPsec tunnel operations due to error
Product-Group=junos
Severity=Major		On all SRX platforms with Internet Protocol Security (IPsec) Virtual Private Network (VPN) tunnel configuration, the srxpfe (SRX Packet Forwarding Engine) process leaks memory during IPsec tunnel operations when memory allocation errors occur. Over time, the increased memory usage grows excessively and eventually causes the srxpfe process to crash, resulting in a temporary IPsec VPN service disruption.
PR NumberSynopsisCategory: Platform infra to support jvision
1928253
Major		The sensord process crashes leading to PFE reboot on MPC10E supported MX platforms
Product-Group=junos
Severity=Major		On MX platforms supporting MPC10E line card, the sensord process crashes due to corrupted memory state on the FPC (Flexible PIC Concentrator). When the sensord process crashes, the PFE (Packet Forwarding Engine) reboots leading to interface flaps and consequent disruption to traffic flow.
PR NumberSynopsisCategory: Layer2 forwarding on EX/NFX/PTX/QFX
1936648
Major		EVPN-VXLAN remote MAC IP programming is removed after a MAC move, resulting in missing ARP on a remote VTEP
Product-Group=junos
Severity=Major		In an Ethernet VPN (EVPN) Virtual Extensible LAN (VXLAN) deployment, a remote MAC move event can cause the remote MAC IP entry to be removed after it is initially installed, resulting in missing Address Resolution Protocol (ARP) state and incomplete hardware programming on the remote Virtual Tunnel Endpoint (VTEP), which can lead to traffic loss for the affected endpoint.
PR NumberSynopsisCategory: Issues related to Junos licensing infrastructure
1925706
Minor		Snmp license polling cause license-check service crash
Product-Group=junos
Severity=Minor		On Junos platforms, a race condition occurs when the Simple Network Management Protocol (SNMP) poller attempts to retrieve device license information while subscriber management and SNMP policing features are in use. This condition causes the license-check process to encounter a null pointer and generate a core dump. There is no traffic or service impact. However, license information retrieval through SNMP polling fail until the license-check service recovers.
PR NumberSynopsisCategory: SW PRs for MPC10E Interfaces
1945343
Major		Disabling one port in the breakout port channel makes another port down
Product-Group=junos
Severity=Major		On all Junos MX platform, when MPC10E interfaces use breakout cabling, administratively disabling one breakout port may unexpectedly impact another port within the same MAC quad due to nondeterministic breakout initialization.
PR NumberSynopsisCategory: FreeBSD Kernel Infrastructure
1932893
Major		vmhost upgrade may fail due to 'Validation failed' when upgrading to freebsd15 based Junos from older releases.
Product-Group=junos
Severity=Major		vmhost upgrade may fail due to 'Validation failed' when upgrading to freebsd15 based Junos from older releases. This does not apply to Junos-EVO.
PR NumberSynopsisCategory: Protocol Independant Multicast
PR NumberSynopsisCategory: Periodic Packet Management Daemon
1931633
Major		PFE crash due to memory corruption when STP is enabled
Product-Group=junos
Severity=Major		On all Junos Platforms, a rare memory corruption condition may occur in the Packet Forwarding Engine (PFE) when Spanning Tree Protocol (STP) is enabled and operating in default (distributed) mode. When the issue is triggered, the PFE crashes and a dc-pfe core file is generated. The exact trigger for the memory corruption is currently unknown.
PR NumberSynopsisCategory: PPPoE functional plugin for bbe-smgd
1939233
Major		Repeated bbe-smgd crashes caused when the system issues a dynamic profile add request during PPPoE session creation, but receives a delete event before the add completes on MX platforms.
Product-Group=junos
Severity=Major		On all MX platforms where Enhanced Subscriber Management (bbesmgd) is enabled, activating PPPoE (PointtoPoint Protocol over Ethernet) over L2TP (Layer 2 Tunneling Protocol) triggers repeated crashes of the bbesmgd process, resulting in core dump generation. This issue affects both Junos and Junos EVO platforms.
PR NumberSynopsisCategory: L2NG PVLAN feature
1916610
Minor		Primary VLAN mapping causes MAC learning failure on PVLAN inner port
Product-Group=junos
Severity=Minor		On EX2300 platform, A Private VLAN (PVLAN) issue occurs where Media Access Control (MAC) learning works correctly for Isolated and Community Virtual Local Area Network (VLAN) access ports, but when the Primary VLAN is assigned to an inner VLAN port, the switch stops forwarding traffic and no Media Access Control addresses are learned after committing the configuration. This issue impacts services on interfaces configured with PVLAN.
PR NumberSynopsisCategory: QFX5K hostpath
1921455
Major		The dcpfe process crashes when adding or removing classifier configuration on QFX5210
Product-Group=junos
Severity=Major		On QFX5210 platform with Class of Service (CoS), the dcpfe process crashes when classifier is configured or removed on interface with active traffic. This can affect traffic forwarding till dcpfe process restarts post crash.
1932314
Minor		Errors "fpc0 ifd null, port X and TD3:ifd null, port X" are seen on QFX5k platforms
Product-Group=junos
Severity=Minor		On QFX5k platforms, if sflow is configured without multicast and BUM traffic is sampled, then the errors "fpc0 ifd null, port X and TD3:ifd null, port X" are observed.
PR NumberSynopsisCategory: for all ipv6 related issues
1922278
Major		IPV6 auto negotiation Router Advertisement packet is missed over configured L2 circuit
Product-Group=junos
Severity=Major		On Junos OS QFX5120 and EX4650, when a L2 circuit is configured between two Customer Premises Equipment (CPE) , IPv6 Router Advertisement packets will not be sent to peer. This will cause devices to not exchange IPv6 messages so they cannot automatically be configured with an IPv6 address, therefore IPv6 connectivity won't be established.
PR NumberSynopsisCategory: QFX L2 PFE
1938291
Major		Deactivating sub-interfaces on flexible ethernet ports leads to VLAN traffic loss
Product-Group=junos
Severity=Major		On Junos platforms having QFX and EX, VLAN(Virtual Local Area Network) traffic loss occur on interfaces configured with flexible-vlan-tagging when an L3 (SP style) sub-interface is deactivated on a port that also carries L2 (enterprise style) VLAN units.
PR NumberSynopsisCategory: QFX L3 data-plane/forwarding
1854995
Major		The dcpfe crash will be seen on Junos QFX5200 platforms due to route churn
Product-Group=junos
Severity=Major		On Junos QFX5200 platforms, a DCPFE (Packet Forwarding Engine Manager) crash may occur due to route churn involving large-scale next-hop changes when any PFE table is near capacity. This crash leads to an FPC restart, causing a complete outage. However, the system will recover automatically.
1905607
Major		Hardware MTU stuck at default value, causing packet drops on VXLAN Interfaces
Product-Group=junos
Severity=Major		On QFX5K and EX4K platforms which are running Junos release, when new VxLAN (Virtual Extensible LAN) vlans with IRBs (Integrated Routing & Bridging) are added, the L3 (Layer 3) interface values overwrite the MTU (Maximum Transmission Unit) entries previously programmed for non-VxLAN vlans that use the same hardware token internally. The VxLAN L3 interface is created with default MTU (1514) because L3 interface MTU value is still 1514 as it is not update by RE (Routing Engine).
1925996
Major		Intermittent uplink ports flaps events observed and syspld read failures on syslog messages were seen.
Product-Group=junos
Severity=Major		Intermittent uplink ports flaps events observed and log messages show continuous syspld read failures during the issue. After a power cycle, the devices recovered.
PR NumberSynopsisCategory: QFX EVPN / VxLAN
1933698
Major		PFE crash due to memory corruption in EVPN-VxLAN type5 overlay ECMP at high tunnel next hop scale
Product-Group=junosvae
Severity=Major		On Junos QFX5110, QFX5120, EX4650, EX4400, EX4100, and EX5200 platforms, when operating with EVPN-VxLAN type5 overlay ECMP at high tunnel next hop scale, the Packet Forwarding Engine (PFE) crash when a memory corruption issue is observed due to a buffer overflow that leads to corruption of heap management structures. This corrupted metadata is detected by the memory manager when later heap allocation or free operations are invoked, resulting in a reported heap corruption condition.
1939298
Major		Traffic drop is seen due to tagged IRB L3 interface with native-vlan and vlan members
Product-Group=junos
Severity=Major		On all Junos OS QFX5K and EX4k platforms, configuring a native VLAN along with VLAN members on an underlay Integrated Routing and Bridging (IRB) Network-to-Network Interface (NNI) that carries VxLAN (Virtual Extensible LAN) traffic results in the VLAN tag not being stripped as expected. Instead of treating the native VLAN traffic as untagged, the interface adds the VLAN tag, leading to packet drops at the remote end.
PR NumberSynopsisCategory: QFX5K JUNOS Interface, MACSec, Optics, SDK, PHY
1938876
Major		Incorrect interface mode leading to 100G Link Flaps on QFX5210-64C
Product-Group=junos
Severity=Major		On Junos OS QFX5210-64C platform, 100G optics inherently operate in CAUI-4 (Chip-to-Module 100 Gb/s Four-Lane Attachment Unit Interface) mode, which is automatically selected by the system and not user-configurable; mismatches due to software versions leading to link instability or flapping.
PR NumberSynopsisCategory: QFX5200/5110/5120/5210 Platfom issues
1938921
Major		Software rollback may cause device to become unresponsive on Junos platforms with qfx-5e image
Product-Group=junosvae
Severity=Major		On Junos platforms running qfx-5e image, the device fails to boot after issuing the command "request system software rollback". Once the issue occurred, the system gets hung and the CLI becomes unresponsive, as a result both control plane and forwarding plane traffic will be impacted.
PR NumberSynopsisCategory: RPD Interfaces related issues
1913519
Major		EVPN routes are stuck in the KRT queue
Product-Group=junos
Severity=Major		When EVPN (Ethernet Virtual Private Network) routes attempt to transition between private (eg, management em1 - with IGP enabled) and public interfaces, it causes an error in next-hop resolution in the kernel, because the system deletes the old indirect next-hop and creates a new one. This happens as the kernel does not support changing an indirect next-hop between private and public interfaces directly.
PR NumberSynopsisCategory: KRT Queue issues within RPD
1929521
Major		Transient multicast traffic loss after Routing Engine switchover when PIM with IGMP or MLD is enabled
Product-Group=junos
Severity=Major		On Junos and Junos OS Evolved platforms where multicast forwarding is enabled, a Routing Engine (RE) switchover, whether planned or unplanned, results in temporary multicast packet loss when active multicast streams are present. This issue is observed only when Protocol Independent Multicast (PIM) is configured together with Internet Group Management Protocol (IGMP) for IPv4 or Multicast Listener Discovery (MLD) for IPv6. During the RE switchover, multicast forwarding state is temporarily removed from the forwarding plane and subsequently restored. While the forwarding state is being rebuilt, multicast packets are dropped for the affected multicast groups. Unicast forwarding is not impacted, and multicast control-plane sessions remain established throughout the event.
1931875
Major		The L2Circuit traffic which are resolving over BGP-LU get drop when 'chained-composite-next-hop' for BGP-LU and 'preserve-nexthop-hierarchy' is configured
Product-Group=junos
Severity=Major		On all Junos and Junos Evolved platforms supporting 'chained-composite-next-hop' for BGP-LU (Border Gateway Protocol - Labeled Unicast), when having L2Circuit (Layer 2 Circuit) links configured, with indirect next-hop resolving on BGP-LU prefix, if 'chained-composite-next-hop' enabled for BGP-LU and with 'preserve-nexthop-hierarchy' configured, L2Circuit traffic will get dropped due to failure in installing the forwarding next-hop with labels for L2Circuit service.
PR NumberSynopsisCategory: Shard routing infrastructure within RPD
1757915
Major		The rpd process crashes when processing multipath routes with mixed indirect and composite next-hops under rib-sharding
Product-Group=junos
Severity=Major		On all Junos and Junos OS Evolved platforms, when rib-sharding is enabled and RT (Route Target) multipath routes containing both indirect and composite next-hop types are processed, the rpd (Routing Protocol Daemon) process will crash due to incorrect handling during the next-hop copy operation from RIB (Routing Information Base) shards to the main RIB thread. An rpd crash results in all routing protocols going down and causes a brief traffic disruption until the rpd process restarts.
PR NumberSynopsisCategory: RPD route tables, resolver, routing instances, static routes
1907558
Major		The rpd process crashes in a vrf having EVPN-VXLAN routes with specific configuration.
Product-Group=junos
Severity=Major		In all Junos and Junos OS Evolved platforms, when EVPN-VXLAN (Ethernet Virtual Private Network-Virtual Extensible LAN) routes are present in a VRF (Virtual Routing and Forwarding), configuring a generate route in same vrf can cause rpd (Routing Protocol Daemon) to crash and restart. Generate route configuration has to be removed to recover from this behaviour.
PR NumberSynopsisCategory: SRX branch platforms
1913911
Major		ARP reply packets may be sent out from the STP blocked port
Product-Group=junos
Severity=Major		On SRX300-series devices, when ethernet-switching is used with spanning-tree protocol enabled, in some cases ARP reply packets may be sent out from a spanning-tree blocked port.
1927646
Minor		Ethernet interfaces with fiber SFPs and configured with 'family ethernet-switching' and 'no auto negotiation' don't come up after reboot
Product-Group=junos
Severity=Minor		After reboot on branch SRX300 Series platforms, interfaces using Small Formfactor Pluggable (SFP) SX/LX fiber transceivers and configured with 'family ethernet-switching' and 'no auto negotiation' might remain in a down state due to incorrect physical medium detection by the Ethernet physical layer (PHY). The issue is timing-dependent, it might happen with single or after several reboots.
1934517
Major		SRX380 XE interface flap on every configuration commit with 1G SFP transceivers
Product-Group=junos
Severity=Major		On the SRX380 platforms, in both cluster and standalone modes, the XE interfaces with 1G SFP fiber transceivers flap once after every configuration commit, causing the device to temporarily lose upstream reachability.
PR NumberSynopsisCategory: SRX COS PFE datapath related issue
1897761
Major		Incorrect reporting of high CPU utilization is observed when there are some packets in the egress queue
Product-Group=junos
Severity=Major		On all SRX platforms, CPU utilization is continuously high when there are some packets in the egress queue. However, CPU was not functionally high and hence no service impact or service degradation is seen.
PR NumberSynopsisCategory: ZT/YT pfe firewall software
1934419
Major		Unexpected throughput reduction seen on AE bundle and physical interface when interface filter and policer are applied together
Product-Group=junos
Severity=Major		The issue is seen on all MX platforms with MPC10/MPC11/LC4800/LC9600 linecards and MX304, when the output filter and interface policer is applied to the egress AE (aggregated ethernet) and physical interface in the output direction and the interface policer runs first then the output filter. This kind of an incorrect policing issue leads to the exhaustion of the policer bandwidth and throughput reduction will be seen.
PR NumberSynopsisCategory: ZT/YTpfe bridging, learning, stp, oam, irb software
1907722
Major		MX/EX92K PFE3 YT Chip Errors Seen During EVPN SLM Multicast Traffic
Product-Group=junos
Severity=Major		PFE error messages are observed on MX304 and EX92K when certain kind of L3 unicast and L3 multicast traffic are transmitted
PR NumberSynopsisCategory: Trio pfe l3 forwarding issues
1927194
Major		When a resolution for an IPv4/IPv6 address fails, NH IFL is throttled causing service impact
Product-Group=junos
Severity=Major		In MX uKern and AFT based cards, NH IFL throttling will be seen when resolution for an destination IPv4/IPv6 address fails.
PR NumberSynopsisCategory: DDos Support on MX
1848317
Major		SCFD flow variation leads to error messages or process crash
Product-Group=junos
Severity=Major		On all Junos MX platforms with line cards MPC10/11/LC9600/LC4800, this includes the MX304 and SCFD (Suspicious Control Flow Detection) enabled, when there are numerous flows being added, deleted, or modified simultaneously, the system experiences error messages or process crashes due to thread synchronization issues. The process crash will cause the FPC to restart.
PR NumberSynopsisCategory: Configuration mgmt, ffp, load-action, commit processing
1562848
Major		The mustd process may crash on all platforms
Product-Group=junos
Severity=Major		With a large-scale configuration, in rare cases, the mustd process might crash. The mustd process, which is responsible for configuration constraint checks, might crash on commit, leading to commit failure.
PR NumberSynopsisCategory: UI Infrastructure - mgd, DAX API, DDL/ODL
1761939
Major		CLI users cannot access configuration mode on Junos and Junos OS Evolved platforms
Product-Group=junos
Severity=Major		On Junos and Junos OS Evolved platforms due to the mgd(Management Daemon) process not releasing commit lock, CLI users cannot change the configuration. CLI users cannot access configuration mode when maximum edit sessions limit is reached in this state. There is no service impact due to this issue.
1821845
Major		The system scripts refresh will fail when using load CLI option
Product-Group=junos
Severity=Major		On all Junos and Junos OS Evolved platforms when the 'edit system scripts' hierarchy configuration is changed using loading a configuration from a file or the terminal using 'load' option, the scripts refresh will fail.
1902358
Minor		RPC crash when non-ASCII character is included in XML data result
Product-Group=junos
Severity=Minor
1947196
Minor		Commitd core dumps observed during specific config commits with "patch generation error - not syncing patch"
Product-Group=junos
Severity=Minor		Resolved an issue where configuration commits could intermittently fail when applying certain routing policy changes. The problem occurred only under specific conditions during commit processing and could result in the commit operation aborting unexpectedly. This fix improves commit stability when updating routing policies.
PR NumberSynopsisCategory: UI Misc
1947442
Major		Deprecate [system syslog shell ] configuration hierarchy
Product-Group=junos
Severity=Major		The [ system syslog shell ] configuration hierarchy is now deprecated and will be removed. All shells that Junos ships will always have syslog commands entered into them as baseline functionality.
PR NumberSynopsisCategory: MX10K linecard
1898825
Major		Unexpected line card restart due to timing condition
Product-Group=junos
Severity=Major		On Junos platforms using line cards LC480 and LC2101, a timing defect in the embedded microkernel thread handling logic causes a panic when a thread attempts to yield execution while interrupt processing is still active. This panic results in a line card reboot.
PR NumberSynopsisCategory: VMHOST platforms software
1927342
Minor		Junos VMhost platforms restart unexpectedly due to deadlock
Product-Group=junosvae
Severity=Minor		On all Junos VMhost platforms, when Junos performs disk access, a rare deadlock involving FreeBSD kernel processes will occur. This condition causes FreeBSD kernel panic, resulting in crash of vmcore and an unexpected device restart, leading to temporary system unavailability.
PR NumberSynopsisCategory: usf ams related issues
1913560
Major		Routing Engine restart on MX platforms with SPC3 line card could cause loss of traffic processing
Product-Group=junos
Severity=Major		On Junos MX960, MX240, and MX480 platforms using SPC3 service line cards, restarting the Routing Engine may result in the network security daemon (nsd) not starting or failing to program internal subsystems like service sets . When this occurs, the control plane becomes unavailable and traffic stops forwarding permanently. This is a timing related behavior observed during the boot sequence and does not appear on every restart.
PR NumberSynopsisCategory: usf flow and datapath issue on SPC3
1925039
Major		Memory leak in ipv4-to-ipv6 session reuse trigger flowd crash
Product-Group=junos
Severity=Major		On Junos OS MX240/MX480/MX960 platforms with MX-SPC3 cards, the flowd process crash and reboots the service PIC when a stale IPv4 session is mistakenly reused for IPv6 due to a memory issue. During process restart the services remain down, session information is briefly lost and active traffic will drop, however the system recovers automatically.

 

 

Extended Solution

25.2R2-S1 - List of Known issues

PR NumberSynopsisCategory: NFX Series Platform Software
1858495
Major		The auto-negotiation is not working properly on NFX350 platform using 1 Gigabit Ethernet SFP
Product-Group=junosvae		On NFX350 platforms connected to certain peer devices over SFP 1 Gigabit Ethernet (GE) with auto-negotiation enabled at both ends, a communication issue occur during the initial connection between devices, causing a mismatch in their status. As a result, the port status on the peer device appear as up/down affecting the traffic.

Resolved In: junos:24.2R2-S2 junos:24.4R2 junos:25.2R1-S1 junos:25.2R2 junos:25.4R1
PR NumberSynopsisCategory: ACX L2 related features
1933501
Major		Device remains down after reboot, when "input-vlan-map tag-protocol-id" is configured but "ethernet-switch-profile tag-protocol-id" is deleted
Product-Group=junos		On all Junos platforms, if an interface is configured with "input-vlan-map tag-protocol-id" and the associated "ethernet-switch-profile tag-protocol-id" is deleted while the VLAN mapping is still active, the system may incorrectly allow the configuration commit. If the device is subsequently rebooted or the chassis is restarted, the device may fail to recover and remain down, resulting in traffic impact.

Resolved In: evo:25.4R2-EVO junos:23.2R2-S7 junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: "agentd" software daemon
1913260
Major		Discrepancy noticed in the interfaces when there is a reset in Linecard or PFE
Product-Group=junos		On all Junos and Junos Evolved platforms, with an established telemetry session with multiple collectors, there will be interface statistics discrepancy when there is a linecard OIR event or PFE offline/online sequence.

Resolved In: evo:22.3X80-D47-EVO evo:22.3X80-D49-EVO evo:23.4R2-S8-EVO evo:26.1R1-EVO junos:23.4R2-S8 junos:24.2R2-S5 junos:24.4R2-S4 junos:25.4R2 junos:26.1R1 junos:26.2R1
1923848
Major		Junos Evolved platfrom, when gNMI collecting data from "optics/lanediags/lane/lane_laser_receiver_power_dbm" and "optics/lanediags/lane/lane_laser_output_power_dbm" are unreadable
Product-Group=junos		The JavaScript Object Notation (JSON) encoding of leafs of type "ieeefloat32"(https://github.com/openconfig/public/blob/master/release/models/types/openconfig-types.yang#L127) is not correct, causing gRPC Network Management Interface (gNMI) data outputs unreadable.

Resolved In: evo:22.3X80-D49-EVO evo:24.4R2-S4-EVO evo:25.4R2-EVO evo:26.2R1-EVO junos:24.2R2-S6 junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: MX YT-ZF Linecards YT, MQSS, Pre-Classifier, HBM Driver Category
1948588
Major		Intermittent fabric errors causing partial or complete traffic loss due to line card PLL lock loss.
Product-Group=junos		On certain MX Series and SRX Series platforms running Junos, an intermittent PLL (Phase-Locked Loop) lock loss on certain line cards can lead to fabric link errors. When the issue happens, forwarding plane traffic across the chassis fabric is impacted, resulting in partial or complete traffic loss.

Resolved In: evo:25.4R2-EVO evo:26.2R1-EVO evo:26.3R1-EVO junos:23.4R2-S1-J9 junos:23.4R2-S7-J13 junos:23.4R2-S8 junos:23.4X14 junos:24.2R2-S2-J18 junos:24.2R2-S2-J5 junos:24.4R2-S3-J17 junos:24.4R2-S4 junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: BBE Layer-2 Bitstream Access
1941118
Major		Modifying "inner-vlan-id-swap-ranges" with active L2BSA subscribers causes new login failures
Product-Group=junos		On Junos MX Broadband Edge (BBE) platforms where Subscriber access management service Layer-2 Bitstream Access (L2BSA) solution is deployed, when "inner-vlan-id-swap-ranges" is modified in the presence of active L2BSA subscribers, free inner-vlan-id count is not calculated correctly, hence subsequent L2BSA login attempts fail and subscribers move to pending state.

Resolved In: evo:26.2R1-EVO evo:26.3R1-EVO junos:23.2R2-S4-J12 junos:23.2R2-S7 junos:26.2R1
PR NumberSynopsisCategory: BBE UP Session Manager related issues
1890895
Major		BNG Controller: Issue with changing the IP address of an active BNG user-plane
Product-Group=junos		BNG Controller: Issue with changing the IP address of an active BNG user-plane

Resolved In:
PR NumberSynopsisCategory: MX304 Routing Engine issues
1642399
Major		Chassis: "KERN_ARP_DUPLICATE_ADDR: duplicate IP address 128.0.0.63! sent from address: 02:01:00:00:00:05 (error count = 1)" seen during bootup
Product-Group=junos		On bootup a harmless error log is seen which does not have any functionality impact and is coming only during bootup due to the static arp configuration of the internal interfaces which in this case seems to be bme0. The static arp configuration are setup for IRI1 and during bootup both the REs tries to assign the IP addresses to their internal interfaces during which this log messages would have come. But this eventually is resolved and internal ip addresses are assigned to the IRI1 interfaces, so one can ignore this error messages for the internal interfaces. The128.0.0.63 is mostly unused in case of Bugatti but present more due to legacy reasons.

Resolved In:
1857029
Major		zeroize command not working
Product-Group=junos		zeroize command not working

Resolved In:
PR NumberSynopsisCategory: MX Platform SW - UI management
1906927
Major		SNMP jnxDomCurrentLaneWarnings OID values are wrong due to other lane values are copied from lane 0
Product-Group=junos		When a interface supports multiple lanes, the SNMP OID jnxDomCurrentLaneWarnings is incorrectly handled as a single lane, resulting in the value from lane 0 being replicated across all other lanes.

Resolved In: junos:23.2R2-S7 junos:23.4R2-S8
PR NumberSynopsisCategory: L2NG Access Security feature
1911142
Major		Client IP assignment failure observed on EX4300 due to DHCPv6 snooping validation issue
Product-Group=junos		On EX4300, DHCPv6 ( Dynamic Host Configuration Protocol for IPv6 ) clients will fail to receive IP addresses due to improper client entry handling when DHCPv6 snooping is enabled.

Resolved In: junos:21.4R3-S12
PR NumberSynopsisCategory: OpenSSL and related subsystems
1939872
Major		xnm-ssl: TLS connection fails with RSA cert and key size of 1024
Product-Group=junos		Starting with stunnel version 5.75, the application operates with OpenSSL Security Level 2 by default when linked with OpenSSL versions that enforce modern security policies. At Security Level 2, OpenSSL requires a minimum RSA key size of 2048 bits. As a result, certificates using RSA 1024-bit keys are no longer accepted. If a service is configured with a certificate using a 1024-bit RSA key, stunnel will fail to load the certificate and the service may fail to start, typically producing an error similar to: SSL_CTX_use_certificate: ee key too small. Impact Systems using 1024-bit RSA certificates for TLS authentication with stunnel may experience service startup failures or TLS handshake failures after upgrading to stunnel 5.75. Recommended Action It is recommended to replace existing 1024-bit RSA certificates with certificates using RSA 2048-bit or stronger keys to comply with modern cryptographic security standards.

Resolved In: evo:25.4R2-EVO evo:26.2R2-EVO evo:26.3R1-EVO junos:25.4R2 junos:26.2R1 junos:26.2R2
PR NumberSynopsisCategory: EVPN ELAN/E-TREE
1901224
Major		Traffic drop seen in EVPN VPWS FXC stream post clear bfd session all.
Product-Group=junos		Traffic drop seen in EVPN VPWS FXC stream post "clear bfd session all" command.

Resolved In: junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: Layer 3 forwarding, both v4+v6
1927951
Major		JUNOS_REG: ACX5448-M: While verifying traffic convergence with FRR on DUT traffic loss duration is more than expected
Product-Group=junos		JUNOS_REG: ACX5448-M: While verifying traffic convergence with FRR on DUT traffic loss duration is more than expected

Resolved In:
PR NumberSynopsisCategory: JUNOS Dynamic Profile Configuration Infrastructure
1930036
Major		High memory usage is showing, when we configure unsupported licensing feature
Product-Group=junos		On all Junos EX and QFX platforms operating in an EVPN (Ethernet Virtual Private Network) environment, this issue occurs when the CLI command 'licensing hourly update' is executed on a device that does not support the licensing feature.

Resolved In: evo:23.4R2-S8-EVO evo:24.2R2-S5-EVO evo:24.4R2-S4-EVO evo:25.4R2-EVO evo:26.2R1-EVO junos:23.2R2-S7 junos:23.4R2-S8 junos:24.2R1-S2-J4 junos:24.2R2-S5 junos:24.4R2-S4 junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: EVO Netstack Juniper Tunnel Driver Module
1919313
Major		The LDP session does not come up when an IPsec SA is configured under OSPFv3
Product-Group=junos		On Junos Evolved platforms, when Open Shortest Path First version 3 (OSPFv3) Internet Protocol Security (IPsec) SA is applied at the interface (IFD/IFL) level, Label Distribution Protocol (LDP) is able to discover the neighbor via User Datagram Protocol (UDP) Hellos, but fails to bring up the Transmission Control Protocol (TCP) session, preventing the LDP session from becoming operational. Disabling the OSPFv3 IPsec SA immediately restores normal LDP operation. During the service impact, the session will be down and traffic will not be send. Once the LDP session comes up, traffic can be routed through it.

Resolved In: evo:23.2R2-S7-EVO evo:23.4R2-S8-EVO evo:24.2R2-S5-EVO evo:24.4R2-S4-EVO evo:25.2R2-S1-EVO evo:25.4R2-EVO evo:26.1R1-EVO evo:26.2R1-EVO
PR NumberSynopsisCategory: EVPN control plane issues
1924472
Minor		EVPN all-active not working in multi-homed setup when router-id not explicitly configured
Product-Group=junos		On Junos OS and Junos OS Evolved platforms, if the router-id is not explicitly configured, the device dynamically selects a router-id during bootup. In this scenario, the device advertises EVPN (Ethernet Virtual Private Network) Type1 AD/ESI (AutoDiscovery/Ethernet Segment Identifier) and Type4 ES RT (Ethernet Segment Route Target) routes with a Route Distinguisher (RD) of 0:0. As a result, Route Targets (RTs) advertised by different Provider Edge (PE) routers end up having identical prefixes. Consequently, only one PE routers RT is advertised and learned. This behavior prevents EVPN all-active redundancy resulting in the loss of multihoming.

Resolved In: evo:26.3R1-EVO
PR NumberSynopsisCategory: EX interfaces issues
1870962
Major		The CPU high utilization is observed after PIC offline/online
Product-Group=junos		On EX4400 platforms with 4x25G or 1x100G ULM (Universal Link Module), the CPU hogs by CMQFX thread for as much as 3.7 secs when Firmware download occurs during PIC offline/online and PFE restart time / Device reboot. To speed up the firmware download operation, the MDIO clock (MDC) frequency is increased from 2.6 MHz to 9 MHz which reduced firmware download time from ~3.8 s to ~2.2 s. This is an enhancement PR.

Resolved In: junos:26.2R1
PR NumberSynopsisCategory: EX optics issues
1887303
Minor		[EX4400-48F] One of the 10gBase-T transceiver is not detected - showing as "Partial" Unknown in PFE
Product-Group=junos		In the EX4400-48F systems, a 10G-BaseT transceiver that was earlier up may not come up post a reboot/image upgrade event; The transceiver may go undetected causing the interface to not be created in the system.

Resolved In:
PR NumberSynopsisCategory: Express PFE L2 fwding Features
1865354
Major		Traffic to anycast IPv6 destination addresses dropped when using ECMP routes
Product-Group=junos		On QFX10002-60C platforms when ECMP (Equal Cost Multi-path) is required to get multiple-path to multiple hosts connected, and with ipv6 address configured as a destination address, the destination MAC rewrite process fails, due to the unilist nexthop for ipv6 destination is getting overwritten. This leads the IPv6 anycast packets from/to the host, to get dropped over the ECMP routes.

Resolved In: junos:22.2R3-S3-J20 junos:22.2R3-S5-J2 junos:22.2R3-S6-J1 junos:22.2R3-S7 junos:23.2R2-S6 junos:23.4R2-S7 junos:24.2R2-S5 junos:24.4R2
PR NumberSynopsisCategory: ACX7332 & ACX7348 Platform Software
1896521
Minor		Intermittent "broadsync_lock_failure" alarms are seen on Junos OS Evolved ACX platforms
Product-Group=junos		On Junos OS Evolved ACX platforms, "broadsync_lock_failure" alarms are seen intermittently without any external trigger. These alarms are seen as a result of internal firmware misconfiguration and affects PTP (Precision Time Protocol) functionality. The alarms are cleared after few seconds (~10 secs) once the broadsync lock is restored.

Resolved In: evo:25.4R2-EVO evo:26.2R1-EVO evo:26.3R1-EVO junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: jdhcpd daemon
1934742
Major		DHCPv6 Relay Active Leasequery not able to establish TCP session to peer
Product-Group=junos		On all MX platforms, when configuring Dual Stack DHCPv6 ALQ without configuring DHCPv4 ALQ (active-leasequery), the ALQ will not be able to establish the TCP (Transmission Control Protocol) session for the ALQ peer due to XID (Exchange Identification) mismatch. This can lead to DHCPv6 subscriber bindings sync issue and server ALQ, resulting in compromising the redundancy functionality.

Resolved In: junos:23.4R2-S7-J21 junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: SRX Datapath Multicast Solution
1935897
Major		Multicast packets are getting dropped at the start of a multicast stream
Product-Group=junos		On SRX345/SRX1500 and SRX1600 platforms , when set security flow strict-packet-order is enabled, a Protocol Independent Multicast (PIM) middle-hop router in a multicast network experience a multicast packet drop at the start of a multicast stream. The packet loss is transient and typically lasts for approximately one second.

Resolved In: junos:24.4R2-S4 junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: Security platform jweb support
1931780
Major		The Captive Web Authentication might not be completed on specific Junos versions for EX Series switch
Product-Group=junos		When a MAC-RADIUS authentication succeeds and the RADIUS server returns CWA redirect attributes (URL-Redirect), the EX switch should intercept client HTTP traffic and return an HTTP 302 redirect to the Captive Portal. The switch instead responded with HTTP 405 (Method Not Allowed), causing the client to bypass the CWA redirect workflow.

Resolved In: evo:25.4R2-EVO evo:26.2R1-EVO junos:23.2R2-S7 junos:23.4R2-S8 junos:24.4R2-S4 junos:24.4X2 junos:25.2R2 junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: Layer 2 Control Module
1930380
Major		The hash collision for storm control profile indices will result in an l2ald process crash
Product-Group=junos		On all Junos OS platforms and Junos OS Evolved platforms which supports storm control, when a different storm-control profile is applied for interface where these profile have same profile index allocated then the storm control profile configuration and system state will not be in sync and a different profile will be applied for interface binding due to profile index collision which results into l2ald process crash.

Resolved In: evo:23.4R2-S8-EVO evo:24.4R2-S4-EVO evo:25.2R2-EVO evo:25.4R2-EVO evo:26.1R1-EVO evo:26.2R1-EVO junos:23.4R2-S8 junos:24.2R2-S5 junos:24.4R2-S4 junos:25.2R2 junos:25.4R1-S2 junos:25.4R2 junos:26.1R1 junos:26.2R1
1938991
Minor		VSTP bridge-priority configuration not taking effect for VLAN group
Product-Group=junos		The VSTP (VLAN Spanning Tree Protocol) bridge-priority configuration does not apply to VLAN (Virtual LAN) groups as expected. This issue results in VLAN groups not reflecting the intended bridge priority. This configuration issue can affect the network's root bridge election process, causing potential inconsistencies in network behavior and performance.

Resolved In: evo:26.1R1-EVO evo:26.2R1-EVO evo:26.3R1-EVO junos:26.1R1 junos:26.2R1
PR NumberSynopsisCategory: Layer2 forwarding on EX/NFX/PTX/QFX
1912050
Minor		Broadcast and Multicast traffic is dropped in MH EVPN MPLS topology where Egress Link Protection is enabled on the PEs and restart l2-learning is executed
Product-Group=junos		On all Junos platform in MH (Multi homed) EVPN (Ethernet Virtual Private Network) MPLS (Multi Protocol Label Switching) topology where Egress Link Protection feature is enabled on the PEs (Provider Edge) routers and the PE - CE (Customer Edge) link is disabled on the one of the MH PEs and 'restart l2-learning' command is executed on another MH PE, this results in broadcast and Multicast traffic drop.

Resolved In: evo:25.2R2-EVO evo:25.4R2-EVO evo:26.2R1-EVO junos:25.2R2 junos:25.4R1-S2 junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: Label Distribution Protocol
1883004
Major		Seeing Degradation in ldp local&remote link failure recovery time with OSPF and ISIS
Product-Group=junos		Dev to add the RN details as informed to Release management team.

Resolved In:
1906611
Major		Crash in the rpd process after LDP P2MP LSP Identifier reaches its maximum value and rolls over, due to duplicate identifier allocation
Product-Group=junos		On all Junos OS and Junos OS Evolved versions that support Label Distribution Protocol Point-to-Multipoint Label Switched Paths (LDP P2MP LSPs), the rpd process (routing protocol daemon) crashes when an LSP Identifier reaches its 24-bit maximum value (224 1 = 16, 777, 215) and rolls over to the starting value because a duplicate identifier is incorrectly allocated. This condition occurs only after prolonged tunnel flapping (typically more than 16 million flaps). When the rpd process crashes, routing convergence is briefly disrupted, and services relying on label-switched traffic are impacted until the process automatically restarts.

Resolved In: evo:26.1R1-EVO junos:26.1R1
PR NumberSynopsisCategory: MPC11 ULC platform software related issues.
1809351
Major		JDI-RCT:M/Mx: ISIS session over MPC11 cards flapped due to "3-Way handshake failed" during ISSU (FRU upgrade stage - reboot phase)
Product-Group=junos		JDI-RCT:M/Mx: ISIS session over MPC11 cards flapped due to "3-Way handshake failed" during ISSU (FRU upgrade stage - reboot phase)

Resolved In:
PR NumberSynopsisCategory: Multicast for L3VPNs
1908581
Major		Significant multicast traffic loss observed during ISSU in Next-Generation Multicast VPN (NGMVPN) deployments
Product-Group=junos		On all Junos platforms that support both InService Software Upgrade (ISSU) and NextGeneration Multicast VPN (NGMVPN), multicast traffic loss can occur during an ISSU operation when NGMVPN deployments use both the Inclusive Provider Multicast Service Interface (IPMSI) and the Selective Provider Multicast Service Interface (SPMSI) with a nonzero threshold. During the ISSU process, Flexible PIC Concentrators (FPCs) upgrade and synchronize, and multicast statistics may briefly report zero. This event causes withdrawal of the SPMSI and a fallback to the IPMSI. If the IPMSI next hops are not fully programmed at that moment, a temporary interruption of SPMSI flows occurs, while IPMSI flows continue forwarding normally. The traffic loss duration is short and typically limited to a few seconds to a few minutes during FPC upgrade cycles, not exceeding the expected ISSU convergence window. Both IPv4 and IPv6 multicast traffic are affected.

Resolved In: evo:26.2R1-EVO junos:26.2R1
PR NumberSynopsisCategory: "ifstate" infrastructure
1863814
Critical		JDI-RCT:AlfaRomeo:: Observing rpd core @ rt_stats_id_grp_set_kernid , krt_stats_id_grp_resp_from_kernel , libsock_build_and_send_rtsock_msg
Product-Group=junos		Release note is added for 25.2R2-S1? RPD core might be seen after GRES. This is a corner case scenario. This issue is not reproduced always.?This is the day-1 behaviour in the rtsock filtering logic. This is a corner case where RPD have filters installed on master and no filters installed on backup. RPD core is seen if gencfg deletes are sent on new master after GRES.??Workaround: RPD restart can help to recover from the issue. The fix is already committed in 25.4DCB and the fix will be committed to 25.2R2-S2.

Resolved In: junos:25.4R1
PR NumberSynopsisCategory: Express Chip L3 software
1827286
Major		The icmpv4/v6 ping fails with ddos-protection* icmp configuration
Product-Group=junos		The PTX10008, PTX10002-60C, or QFX10002-60C platforms may not send back ICMPv4/v6 reply packets properly due to defects leading to misprogramming of hardware. Ping with v4/v6 from another device to the PTX10008, PTX10002-60C, or QFX10002-60C platform will fail.

Resolved In: junos:22.4R3-S5 junos:22.4X50
1877538
Major		Multicast traffic loss is seen when MVPN with node protection is enabled
Product-Group=junos		On Junos PTX and QFX10K platforms with node protection enabled on a Multicast Virtual Private Network (MVPN) scenario, multicast traffic loss will be seen when the number of child links in an aggregated ethernet (AE) interface for bypass Label Switched Path (LSP) egress interface is higher than primary LSP and one of the child links goes down on primary LSP egress interface.

Resolved In: junos:21.4R3-S12 junos:22.4R3-S8 junos:23.2R2-S5 junos:23.4R2-S6 junos:23.4R2-S7 junos:24.2R2-S2 junos:24.4R2
PR NumberSynopsisCategory: Issues related to PKI daemon
1919729
Major		SRX 2300 SRX 4700 SRX4300 and SRX1600 Enter System Halt or Amnesiac State After Reboot When TPM/MEK is Enabled
Product-Group=junos		On SRX4300, SRX 4700, SRX2300 , SRX1600 platforms When Master password and Master Encryption Key is set on the device and rebooted , the system was getting into amnesiac state.

Resolved In: junos:24.2R2-S5 junos:24.4R2-S4 junos:26.1R1 junos:26.2R1
PR NumberSynopsisCategory: QFX5200/5110/5120/5210 Platfom issues
1725116
Major		Sometimes 100G link will go down and will remain down
Product-Group=junosvae		On the Junos QFX5200 platform, sometimes 100G link will go down and will remain down

Resolved In: junos:20.3X75-D43 junos:20.4R3-S8 junos:21.2R3-S6 junos:21.3R3-S5 junos:21.4R3-S4 junos:22.1R3-S3 junos:22.2R3-S1 junos:22.3R3
1850037
Major		Traffic loss or random reboots on QFX5K platforms due to host file system corruption
Product-Group=junosvae		On Junos QFX5K platforms running qfx-5e images, ungraceful shutdown may lead to file system corruption. Corruption of the RPM database in the Host OS can lead to several issues, including timeouts in certain Junos CLI show commands such as "show version", problems with FPCs (Flexible PIC Concentrators) or other FRUs (Field Replaceable Units) like power supplies and fan trays failing to initialize, or the device undergoing unexpected watchdog reboots.

Resolved In: junos:22.2R3-S6 junos:23.4R2-S4 junos:24.4R2 junos:25.1R1 junos:25.2R1
PR NumberSynopsisCategory: RPD API infrastructure
1905326
Major		JDI-REG : VIRTUAL : JUNOS : While verifying the Prpd Client on r0 programmable-rpd clients details of 'SwanAgent' are not obtained as expected.
Product-Group=junos		JDI-REG : VIRTUAL : JUNOS : While verifying the Prpd Client on r0 programmable-rpd clients details of 'SwanAgent' are not obtained as expected.

Resolved In:
PR NumberSynopsisCategory: SW PRs for SCBE3 fabric
1911536
Minor		Ungraceful removal of SCB causes traffic loss
Product-Group=junos		On Junos OS, MX Series (MX240, MX480, MX960) devices that use a Switch Control Board (SCB), forwarding traffic may be interrupted for approximately one minute if an SCB is physically removed without first powering it off. The control plane remains operational, but forwarding is impacted until fabric manager and forwarding services selfrecover.

Resolved In: evo:25.4R2-EVO evo:26.1R1-EVO evo:26.2R1-EVO junos:25.4R1-S2 junos:25.4R2 junos:26.1R1 junos:26.2R1
PR NumberSynopsisCategory: Generic platform and infra issues for MS-MIC and MS-MPC(XLP)
1899178
Critical		Service session drops are observed when CPU throttling is configured on platforms with service cards installed
Product-Group=junos		On all Junos MX platforms that have MS-MPC or MS-MIC service cards installed, the use of the CPU throttling can cause the production service sessions to be dropped.

Resolved In: junos:21.2R3-S10 junos:21.2R3-S6-J16 junos:22.4R3-S7-J5 junos:22.4R3-S9
PR NumberSynopsisCategory: ZT/YT pfe l3 forwarding issues
1886395
Major		FPC crash is seen on Junos platforms in a rare scenario
Product-Group=junos		On all Junos platforms, FPC (Flexible PIC Concentrator) crashes due to panic caused by incorrect handling of an application. This causes service impact since the card restarts after crash.

Resolved In: evo:25.2R2-EVO evo:25.4R1-EVO junos:23.4R2-S4-J23 junos:23.4R2-S4-J30 junos:23.4R2-S6 junos:24.2R2-S2-J16 junos:24.2R2-S4 junos:25.2R2 junos:25.4R2
PR NumberSynopsisCategory: Issues related to broadband edge apps (PPP, DHCP) on Trio ch
1924812
Minor		Dual stack PPPoE/DHCPv6 syslog reports PFE_ERROR_INVALID_STATE fd01:3333:b205:44f1:: /64 => source lookup is not enabled for flow 18558885
Product-Group=junos		Dual stack PPPoE/DHCPv6 syslog reports PFE_ERROR_INVALID_STATE fd01:3333:b205:44f1:: /64 => source lookup is not enabled for flow 18558885

Resolved In: evo:25.4R2-EVO evo:26.2R1-EVO evo:26.3R1-EVO junos:25.4R2 junos:26.2R1
1929254
Major		FPC crash triggered by peer member restart in AE bundle
Product-Group=junos		On Junos MX platforms with MPC1-MPC9 line cards, subscriber over AE (Aggregated Ethernet) is enabled, one of the AE member link's line cards (FPC) restarts, causes an unexpected crash on another member links line card. When the crash occur, PFEMAN is stuck.

Resolved In: junos:25.4R2 junos:26.2R1
PR NumberSynopsisCategory: Trio pfe l3 forwarding issues
1922741
Minor		FPC heap memory will be leaked when CCNHs are recreated due to VPLS PNH are getting deleted and re-added
Product-Group=junos		On Junos MX platforms with MPC1-9/LC480/LC2101 (including MX204 and MX10003 platforms) when "set protocols l2circuit resolution preserve-nexthop-hierarchy" is enabled, FPC heap memory will leak when CCNHs (Chained Composite Next Hops) are recreated because of the VPLS (Virtual Private LAN Service) PNH (Preserve Nexthop Hierarchy) are getting deleted and re-added.

Resolved In: junos:22.4X4 junos:23.2R2-S7 junos:23.4R2-S8 junos:24.2R2-S4 junos:24.4R2-S3 junos:25.2R2 junos:25.4R2 junos:26.1R1 junos:26.2R1
PR NumberSynopsisCategory: DDos Support on MX
1897237
Major		Traffic flow display not accurate when SCFD is enabled
Product-Group=junos		On MX platforms with MPC10/MPC11/LC9600/LC4800 linecards and MX304/MX301 platforms, if SCFD (Suspicious Control Flow Detection) is enabled and lot of flow are tracked on the device, error logs may be reported when the table overflows. This is purely a display issue.

Resolved In: evo:25.2R2-EVO evo:25.4R1-EVO junos:23.4R2-S4-J36 junos:23.4R2-S8 junos:25.4R1
PR NumberSynopsisCategory: Ephemeral Database
1717477
Major		The system processes accessing the ephemeral database can crash
Product-Group=junos		On all Junos and Junos OS Evolved supporting ephemeral databases, due to a race condition, when a system process reads configuration from ephemeral database and in parallel, there is a commit in the static database, the system process crashes.

Resolved In: evo:22.2R3-S4-EVO evo:22.3R3-S3-EVO evo:22.3X50-EVO evo:22.3X80-D49-EVO evo:22.4R0-J0-EVO evo:22.4R3-EVO evo:23.2R1-S2-EVO evo:23.2R2-EVO evo:23.3R1-EVO junos:21.2R3-S5 junos:21.2X32-D20 junos:21.2X32-D30 junos:21.2X33 junos:21.2X9 junos:22.2R3-S4 junos:22.3R3-S3 junos:22.3X60 junos:22.4R3 junos:23.2R1-S2 junos:23.2R2 junos:23.3R1
PR NumberSynopsisCategory: UI Infrastructure - mgd, DAX API, DDL/ODL
1735584
Minor		Execution of get-directory-usage-information RPC on yang based client or controller fails with validation error
Product-Group=junos		Execution of get-directory-usage-information RPC on yang based client or controller fails with validation error

Resolved In:
1784818
Major		The non-root user will not be able to copy files
Product-Group=junos		On all Junos and Junos Evolved platforms, when logged in as a non-root user and trying to copy a file from a remote location, it shows as cannot become non-root username although logged in as a non-root user and an error message is thrown.

Resolved In: evo:21.4R3-S9-EVO evo:21.4X9-EVO evo:22.2R3-S6-EVO evo:22.3R3-S4-EVO evo:22.3X50-EVO evo:22.3X80-D43-EVO evo:22.3X80-D45-EVO evo:22.4R0-J0-EVO evo:22.4R3-S4-EVO evo:23.2R2-EVO evo:23.4R1-S1-EVO evo:23.4R1-S2-EVO evo:23.4R2-EVO evo:24.2R1-EVO evo:24.2R2-EVO junos:20.2R3-S10 junos:20.2X42 junos:20.3X75-D36 junos:20.3X75-D441 junos:21.2R3-S9 junos:21.2X32-D30 junos:21.2X33 junos:21.2X34 junos:21.2X35 junos:21.2X40 junos:21.2X9 junos:21.4R3-S9 junos:22.2R3-S6 junos:22.3R3-S4 junos:22.3X60 junos:22.4R3-S4 junos:22.4X4 junos:22.4X50 junos:23.2R2 junos:23.4R1-S1 junos:23.4R2 junos:24.1R1 junos:24.2R1 junos:24.2R2 junos:24.3R1 junos:24.3R2 junos:24.4R1 junos:24.4R2
PR NumberSynopsisCategory: VSRX platform software
1922165
Major		ED25519 ssh key now supported from AWS and GCP upon vSRX 3.0 instance creation
Product-Group=junos		Starting in 25.4R2, customers can use ed25519 ssh keys from either the web console or CLI of AWS and GCP as well as rsa ssh keys when selecting an SSH key to be able to log into the system with at launch. Configuring the keys through cloud-init config was already supported.

Resolved In: junos:25.4R2 junos:26.1R1 junos:26.2R1
1945668
Minor		Upgrading vSRX to Junos 24.4R2-S3 and above causes AE bundling issues and traffic loss
Product-Group=junos		Upon booting up a Junos vSRX platform following an upgrade to Junos 24.4R2-S3 and above, one or more interfaces may not be seen within the interfaces list from running a "show interfaces terse" for aggregate ethernet (AE) interfaces. Due to this it causes AE bundling issue and interfaces don't come up leading to traffic loss.

Resolved In: junos:25.4R2 junos:26.2R1

 

Modification History

First publication 2026-06-01