21.4R3-S3: Software Release Notification for JUNOS Software

Alert Type

SRN - Software Release Notification

RiskRisk Description

Low/Notification

Software Release Notification

ImpactImpact Description

Low/Notification

Software Release Notification

Product Affected

ACX EX MX NFX PTX QFX SRX vSRX

Alert Description

Junos Software Service Release version 21.4R3-S3 is now available for download from the Junos software download site

Download Junos Software Service Release:

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

Solution

Junos Software service Release version 21.4R3-S3 is now available.

21.4R3-S3 - List of Fixed issues

PR Number	Synopsis	Category: L2NG SFLOW feature
1699585	Adaptive sampling will not work if the system clock is turned backward Product-Group=junos	On Junos platforms, if the system clock is turned backward then adaptive sampling will not work.
PR Number	Synopsis	Category: EX4300 PFE
1697685	Gratuitous ARP reply will not update the ARP cache Product-Group=junos	On EX4300 platforms, the ARP (Address Resolution Protocol) cache will not be updated upon receiving the Gratuitous ARP reply even with "gratuitous-arp-reply" configured.
PR Number	Synopsis	Category: EX4300 Platform
1678506	The interface on the device will go down when one or more interfaces are connected to the Advantech3260 device at another end Product-Group=junosvae	On EX4300 platforms with version 19.1R3 and above, the interface on the device will go down when the device is connected to the Advantech3260 device at the other end. It will have a traffic impact.
PR Number	Synopsis	Category: EX4300 Filters implementation
1699777	TCAM space might be exhausted when learning DHCP snooping entries on a trusted port Product-Group=junos	On EX, QFX5k, and MX platforms having persistent binding for DHCP (Dynamic Host Configuration Protocol) snooping configured might cause TCAM (Ternary Content Addressable Memory) space exhaustion for DHCP snooping learning on the trusted port after the device reboot.
PR Number	Synopsis	Category: EX4300 Layer 2 implementation
1697995	Dot1x authentication failure for EVPN VXLAN enabled port Product-Group=junos	On Junos EX4300-48MP platforms, dot1x authentication fails with EVPN-VXLAN (Ethernet VPN-Virtual extensible LANs) end to end configuration on a user port.
PR Number	Synopsis	Category: EX2300/3400 PFE
1685938	MAC address learning might not happen on specific EX/QFX platforms Product-Group=junos	On specific EX/QFX platforms (EX2300/EX3400/EX4300-MP/EX4100/EX4400/QFX5K), MAC learning is not happening in any of the interfaces of the particular VLAN (contains both LAG (Link Aggregation Group) and non-LAG interfaces) when VSTP (VLAN Spanning Tree Protocol) is enabled on LAG interface with no child interface and VSTP is disabled for non-LAG interfaces. It causes a traffic impact as mac address learning does not happen for all interfaces on the VLAN.
1706845	Layer 3 forwarding issues for IRB Product-Group=junos	On EX2300, EX3400, EX4400, EX4100 and EX4300-MP platforms, when master FPC with mac-persistence-timer configured on a virtual chassis switch is halted, it leads to layer 3 (l3) forwarding issues for the integrated routing and bridging (IRB).
PR Number	Synopsis	Category: EX2300/3400 platform
1692579	Few uplink ports of EX2300-48MP are not coming up Product-Group=junos	Connecting 1G transceiver to port 4 or 5 (ge-0/1/4 and ge-0/1/5) of PIC 1 (Physical Interface Cards) of EX2300-48MP causes link down.
PR Number	Synopsis	Category: EX-Series VC Infrastructure
1689946	Instability observed after mastership switchover on members with SFP-T pluggable installed on EX4600-VC Product-Group=junos	On Junos EX4600 Virtual Chassis (VC), a routing engine master switchover may lead members to disconnect from the VC for approx. 2 mins before the members re-join the VC. This instability will lead to traffic loss. This only happens on members with SFP-T pluggable installed.
PR Number	Synopsis	Category: SPC3 HW and SW Issues
1692611	SRX cluster may fail in a rare scenario when node status changes to disabled state without going through the ineligible state Product-Group=junos	On SRX1500/SRX4100/4200/SRX4600/SRX5400/5600/5800 with chassis cluster configured, the RE(Routing-Engine) & PFE (Packet Forwarding Engine) connection may break, and the cluster may fail when the node status changes from primary/secondary state to disabled state without going through the ineligible state.
PR Number	Synopsis	Category: CoS support on ACX
1689604	EVPN Packets may go to incorrect queues due to the wrong classification and may lead to packets drop during congestion Product-Group=junos	On Junos ACX5448 and ACX710 devices, in EVPN (Ethernet Virtual private network) with a multihoming scenario, if one link fails traffic will switch over to another link. When FPC is restarted and the dot1p classifier is changed, restore the failed link. Packets may not go to the correct queue due to the wrong classification when traffic switches back to the original link, and traffic drops may happen during congestion.
PR Number	Synopsis	Category: ACX Services feature
1633395	Delegated BFD sessions configured on routing-instance may fail to come up Product-Group=junos	On ACX5448 and ACX710 platforms, all types of delegated BFD sessions configured on routing-instance other than the default routing-instance might not come up.
PR Number	Synopsis	Category: "agentd" software daemon
1668319	GRPC Server not decoding leaf-list correctly Product-Group=junos	In cases where scalar arrays are the last leaf elements in a container, due to the missing pop of the key of the scalar array element, the path stack keeps growing up.
PR Number	Synopsis	Category: Category for tracking AMR related issues.
1643570	AMR first session traffic is not copying over multiple paths for v6 traffic over v6 ipsec tunnel mode Product-Group=junos	AMR when it is enabled in non-cso v6 over v6 mode with IPSEC tunnels, the first session after reboot or forward restart, will not have multipath treatment, post that the feature works fine.
PR Number	Synopsis	Category: a20a40 specific issue
1698797	Fabric monitoring suspension and control link failure may cause HA cluster outage Product-Group=junos	A series of event happens to trigger this issue: I. HW failure on Node0 which causes a major chassis alarm to suspend fabric link monitoring II. A Critical HW failure happening again on Node0 which also causes control link down. III. Due to above, Node1 is neither able to see Node0 going into disabled state, not fab link down as which was suspended earlier in step I.
PR Number	Synopsis	Category: australia related kernel issue
1714002	vmcores can be seen on SRX5k platforms when the fxp0 interface is configured under management-instance Product-Group=junos	On Junos SRX5K series platforms, when the 'set system management-instance' command is committed on the secondary node, continuous VMcores are observed on the secondary node leading to the kernel crash. The secondary node of the cluster reboots automatically to recover from the error. The cluster redundancy is not restored until management-instance knob is removed using "delete system management-instance".
PR Number	Synopsis	Category: srx5k service offloading related PR
1702138	TCP packet drops are seen when services-offload is enabled Product-Group=junos	On Junos SRX5400, SRX5600, SRX5800 and SRX4600 Platforms, packet drops may be observed while services-offload is enabled.
PR Number	Synopsis	Category: Bi Directional Forwarding Detection (BFD)
1706018	The BFD session would flap when the GRES is triggered with single-hop BFD over AE interfaces configured Product-Group=junos	On all Junos and Junos Evolved platforms, Bidirectional Forwarding Detection (BFD) sessions would flap and leads to route churn in the network when the Graceful Routing Engine Switchover (GRES) is triggered with single-hop BFD sessions over Aggregated Ethernet (AE) interfaces. It is a timing issue.
PR Number	Synopsis	Category: Border Gateway Protocol
1652666	Wrong next-hop weight might be observed with BGP PIC enabled Product-Group=junos	On all Junos and Junos OS Evolved platforms, wrong weight might be observed with BGP PIC (Border Gateway Protocol Prefix Independent Coverage) enabled.
1659441	Routing Process Daemon (rpd) crashes and restarts when a specific timing condition is hit with BGP configuration Product-Group=junos	On all Junos platforms and all Junos Evolved platforms, Routing Process Daemon (rpd) crashes and restarts when BGP (Border Gateway Protocol) is configured and a specific timing condition is hit for secondary route. This issue might cause a traffic impact.
1662199	The rpd will crash when using BGP-CT Product-Group=junos	On all Junos and Junos OS Evolved platforms, the rpd will be crashed under Border Gateway Protocol Classful Transport (BGP-CT){SR-MPLS} scenario when any route is added/modified/deleted multiple times.
1679646	The AGGREGATOR attribute will not be set correctly when the independent-domain is configured Product-Group=junos	On all Junos and Junos OS Evolved platforms, in BGP(Border Gateway Protocol) deployments with the "independent-domain" and aggregate routes configured under routing-instance, the AGGREGATOR attribute encapsulated within the ATTR_SET will not be populated correctly. This can result in unintended behavior or service impact because the BGP neighbors receiving this update may discard it as it is not RFC complaint.
PR Number	Synopsis	Category: BBE Remote Access Server
1697392	A few subscriber sessions will not be up post RE switchover Product-Group=junos	On MX platforms, in a high scale subscriber scenario with close to 100% address pool utilization, when the session table and the IP pool database are out of sync and a switchover is performed, authd assigns an in-used IP address to the new subscriber session. This issue doesn't have an impact on existing subscribers but is not able to keep the new sessions up.
1710145	High CPU utilization is seen on Junos MX series platforms Product-Group=junos	This issue is seen on Junos MX series platforms that support authd process. when authd traces are enabled in scale setup with 10k+ routing instances, authd will keep generating some logs. These logs consumes CPU cycles and makes authd process to utilize 100% CPU.
PR Number	Synopsis	Category: Express Broadway PFE L3
1697827	Traffic drop is observed after deleting or deactivating the logical interface Product-Group=junos	On Junos QFX10K platforms configured with multiple logical interfaces (IFLs) on the same physical interface (IFD), if any one of the logical interfaces (IFL) is deleted or deactivated, it will cause a complete traffic drop on the other IFLs of the same IFD.
PR Number	Synopsis	Category: Virtual-chassis platform/chassisd infrastructure PRs for MX
1670345	The PFE I/O chip setup failed for some interfaces and causes those interfaces missing in PFE after backup chassis upgraded via Sequential Upgrade Product-Group=junos	When the MX virtual-chassis was upgraded by using the Sequential Upgrade method, there is the possibility that PFE provisioning may start before link training completes and all PICs are online. In such scenario, the ifd provisioning is preserved and if the preserved state is applied to the PFE before fabric training has completed and all the PICs have been powered on, ifd missing errors will be seen.
1697630	MX-VC: The backup VC router could become master after the system reboot Product-Group=junos	On Junos MX platforms when VC (Virtual Chassis) configuration is done and after the VC membership election, when routers are rebooted, the previously elected primary router will become the backup router and the backup router will come up as the primary router.
PR Number	Synopsis	Category: Class of Service
1693977	The oid tree jnxCosQstatEntry returns nothing for some interfaces after restarting class-of-service Product-Group=junos	The oid tree jnxCosQstatEntry returns nothing for some interfaces after restarting class-of-service.
1699714	The AE (Aggregated Ethernet) link flaps on MX platforms with MPC10, MPC11, and LC9600 when high or medium priorities are configured on the queue Product-Group=junos	The AE link flaps on MX platforms with MPC10, MPC11, and LC9600 when high or medium priorities are configured on the queue. When there is a link flap for a link in the AE bundle with queues having configured high/medium priorities, then it is possible to see a queue drain failure which leads to the queue getting stuck with no traffic moving out.
PR Number	Synopsis	Category: Enhanced Broadband Edge support for cos
1684019	Traffic would hit wrong queue post ISSU Product-Group=junos	On all MX platforms configured with Class-of-Service scheduler maps, after In-Service Software Upgrade (ISSU) the scheduler map does not have the right queue mapping and the traffic would be mapped to the wrong queue leading to traffic loss.
1687642	CoS memory errors are seen when "chassis traffic-manager enhanced-priority-mode" is configured Product-Group=junos	On MX platforms, when "chassis traffic-manager enhanced-priority-mode" is configured, Class of Service (COS) related memory errors are seen. The initial error "Uninitialized Memory" has no impact, but in a rare case with a subscriber high scale, COS resources might be exceeded and report COS_HALP allocation errors which would require an FPC reboot.
PR Number	Synopsis	Category: QFX Access Control related
1705490	EAP authentication might not be successful with 802.1X server-fail configuration Product-Group=junos	On all Junos and Junos OS Evolved platforms, in the case of server-fail scenarios for Extensible Authentication Protocol (EAP) based authentication, the supplicant will be shown as authenticated but the supplicant failed to get authenticated. This will impact the users not being authenticated even if the server-fail fallback option is configured as permit.
PR Number	Synopsis	Category: QFX Control Plane VXLAN
1712805	The generation of the VXLAN table appears to be lost after loading configuration Product-Group=junos	On all Junos and Juos OS Evolved platforms that have EVPN-VXLAN enabled, the generation of the VXLAN table and the message during the setup of the EVPN/VXLAN connection seems to get lost because of a timing issue between the layer2 address learning daemon(l2ald) and routing protocol daemon( rpd ) connections.
PR Number	Synopsis	Category: Device Configuration Daemon
1658016	Observed error log "UI_CONFIGURATION_ERROR Product-Group=junos	Due to the issue, an error log is printed, and the dcd is restarted. But there is no functionality impact for BFD sessions. There may be a slight delay in the new config taking effect as dcd is restarted.
1695663	The Unicast traffic is dropped on QFX5100/EX46xx-VC platforms Product-Group=junos	On QFX5100/EX46xx Junos platforms configured with Virtual Chassis(VC), if a master member is unplugged or forced to power off, the unicast traffic is dropped due to mac-persistence-timer expiry there is a difference in mac addresses between logical aggregated parent interface and member aggregated ethernet(ae) interface.
PR Number	Synopsis	Category: Firewall Filter
1697959	Deactivating and activating the GRES causes churn in dfwd filter addition/deletion Product-Group=junos	On all Junos dual-RE platforms, when performing activate/deactivate Graceful Routing Engine Switchover (GRES) multiple times synchronization issues are observed between the master and backup dfwd process.
PR Number	Synopsis	Category: ACX BFD specific issues
1670684	New BFD sessions will not come up on ACX5448/ACX710 due to continuous flaps Product-Group=junos	On ACX710/ACX5448 platforms, if the Bidirectional Forwarding Detection(BFD) endpoint ID in chipset level reaches ~8K (for ACX710) or ~16k (for ACX5448), the following BFD session can't be created properly and be stuck at Init state. Once the issue is hit, disable BFD configuration and restart the bfdd process to make the endpoint ID roll over from 16.
PR Number	Synopsis	Category: CoS support on DNX
1703840	CoS rewrite rules will not work in L3VPN scenario Product-Group=junos	On Junos OS ACX710/ACX5448 platforms, in Layer 3 Virtual Private Network (L3VPN) scenario, Class of Service (CoS) Experimental (EXP) rewrite rules configuration will not take effect. Thus CoS remarking will not work properly leading to the wrong remarking on Multiprotocol Label Switching (MPLS) packets.
PR Number	Synopsis	Category: Gnats category for dynamic rendering infrastructure
1695523	The routing protocol daemon may core dump when streaming telemetry data Product-Group=junos	There are edge cases where the routing protocol daemon may core dump when streaming the telemetry data for paths under the protocol container. For example: "/network-instances/network-instance[name='DEFAULT']/protocols/protocol/isis"
PR Number	Synopsis	Category: Control Plane for Node Virtualization
1683312	'clear interfaces statistics all' taking more than 9 min due to invalid PIC configuration inside GNF. Product-Group=junos	Invalid PIC configuration inside GNF may add delay to 'clear interface statistic all' command. This issue does not impact any functionality. chassis fpc pic
PR Number	Synopsis	Category: EVO Class of Services
1692542	The host outbound traffic drop is seen on all Junos OS Evolved platforms due to a race condition between class-of-service host-outbound-traffic and forwarding-class configuration update Product-Group=junos	Host outbound forwarding-class/queue mapping does not take effect as expected per configurations on all Junos OS Evolved platforms. Host outbound queue is not getting correctly programmed and caused a traffic drop.
PR Number	Synopsis	Category: EVO L2 Control Plane PRs
1703226	The l2ld process will crash when an IFL is changed to trunk mode and a new VLAN is added Product-Group=junos	On all Junos Evolved platforms, the Layer 2 address learning daemon (l2ald) process will crash when there is a scaled configuration to commit and a Logical Interface (IFL) mode is changed from access to the trunk and a new Virtual Local Area Network (VLAN) is added to the IFL in a single commit. This issue will impact new MAC learning until the l2ald process recovers.
PR Number	Synopsis	Category: AAA, auditd issues
1622936	Idle/inactive netconf user sessions not terminated as per the config set under class Product-Group=junos	Session idle time was not supported for netconf session in EVO, as part of the current PR extending this feature support. This might result in dangling sessions.
PR Number	Synopsis	Category: EVPN Layer-2 Forwarding
1662954	In EVPN-MPLS Mutihoming scenario DF election will get stuck in the Preference based state Product-Group=junos	On all Junos and Junos Evolved platforms supporting preference-based DF (Designated Forwarder) election in EVPN-MPLS Multihoming scenario, the DF election will be stuck in the Preference based state if there will be a change in the df-election on the fly.
1702016	ARP/ND doesn't resolve when extended-vlan-list is configured for the specific VLAN Product-Group=junos	On all Junos & Junos Evolved platforms, ARP(Address resolution protocol)/ND(Neighbour discovery) doesn't resolve when extended-vlan-list is configured for the specific VLAN(virtual local area network).
1709007	In EVPN scenario, proxy-arp on IRB interfaces do not work as expected Product-Group=junos	On all Junos and Junos Evolved platforms with EVPN(Ethernet Virtual Private Network) scenario, IRB(Integrated Routing and Bridging) interfaces send proxy-arp for IPs that are already present in the network, which triggers a duplicate IP error on the end device/host.
PR Number	Synopsis	Category: Control plane EVPN multicast
1670435	EVPN multicast traffic may get impacted because of routes getting stuck in the kernel routing table (krt) queue Product-Group=junos	On all Junos and Junos Evolved platforms with Ethernet Virtual Private Network(EVPN) enabled, in the presence of certain triggers like rpd restart, rollback of configuration etc., EVPN multicast routes might get stuck in the kernel routing table (krt) queue due to which the corresponding routes in the forwarding table may point to 'fictitious' next hops, resulting in traffic disruption.
PR Number	Synopsis	Category: EX4400 PFE software
1694800	On a PVLAN with DAI ARP packets will be forwarded between isolated ports Product-Group=junos	On Junos based EX platforms, when Dynamic ARP (Address Resolution Protocol) Inspection (DAI) is configured on PVLAN (Private Virtual Local Area Network), ARP packets coming from the LAG (Link Aggregation Group) are forwarded to other isolated access ports.
1704470	Traffic drops observed with hierarchal overlay ECMP configuration Product-Group=junos	On EX4400 platforms in the EVPN-VXLAN environment, the overlay ECMP (equal-cost multipath) route does not get programmed in hardware resulting in traffic drops when hierarchal overlay ECMP is configured.
PR Number	Synopsis	Category: EX4400 platform
1683753	EX/QFX SNMP: jnxOperatingDescr.1.1.0.0 returns blank, but jnxOperatingState.1.1.0.0 returns value. Product-Group=junos	On EX/QFX series switch, jnxOperatingDescr.1.1.0.0 returns blank, but jnxOperatingState.1.1.0.0 returns value.
1706952	The PoE firmware upgrade fails on EX4400 platforms Product-Group=junos	On Junos EX4400 platforms, the Power over Ethernet (PoE) firmware (FW) upgrade fails and gets stuck in "POE_SW_ERASE" state.
PR Number	Synopsis	Category: EX Entry Level Access VC platform
1704457	The PXE BIOS recovery fails on EX9204/9208/9214 VC setup Product-Group=junos	On EX9204/9208/9214 platforms configured in a Virtual Chassis(VC) setup, if the SSD(solid-state drive) local disk is erased or there is a hardware failure for EX9200-40XS FPC, the EX9200-40XS recovery will fail if performed through the PXE(Preboot Execution Environment) and stuck in a boot loop.
PR Number	Synopsis	Category: Express PFE including evpn, vxlan
1711889	Traffic drop is observed in the EVPN-VXLAN scenario with Type-2 ESI Tunnel Product-Group=junos	On Junos QFX10K platforms, in an Ethernet VPN-Virtual Extensible LAN (EVPN-VXLAN) scenario, if the route is learned over ECMP (Equal-Cost Multi-Path) and each child next hop of the ECMP group is learned over Type-2 Ethernet Segment Identifier (ESI) tunnel, the route doesn't get programmed correctly on the device, causing the complete traffic drop for the affected route.
PR Number	Synopsis	Category: Express PFE L2 fwding Features
1693424	Traffic loss is observed when the ECMP path is IRB over AE (IPv4 -> MPLS) Product-Group=junos	On QFX10002 switches, when the ECMP path is integrated routing and bridging (IRB) over aggregated Ethernet {IPv4 -> MPLS}, this results in packets getting discarded and traffic loss is observed.
1709605	Traffic drop is observed with "vxlan encapsulate-inner-vlan" knob on Junos QFX10K platforms Product-Group=junos	On Junos QFX10K platforms, in an EVPN-VXLAN (Ethernet VPN-Virtual Extensible LAN) scenario, if the "vxlan encapsulate-inner-vlan" knob is enabled, the Address Resolution Protocol (ARP) unicast packets are dropped if it egresses out of the VXLAN tunnel. This will cause a complete traffic drop.
PR Number	Synopsis	Category: Express PFE MPLS Features
1712076	FPC memory leak will cause FPC crash Product-Group=junos	On PTX and QFX10K platforms, enabling collection of LSP (label-switched path) statistics for telemetry data will cause FPC (Flexible Physical Interface Cards (PIC) Concentrators) memory leak resulting in FPC crash.
PR Number	Synopsis	Category: SRX1500 platform software
1690621	SRX1500 chassis cluster port ge-0/0/1 does not work in switching mode Product-Group=junos	SRX1500 chassis cluster port ge-0/0/1 on a node is not working in switching mode when this interface is configured in L2 (layer2) mode.
PR Number	Synopsis	Category: GMIC2 platform driver issues
1693211	MACsec on logical interfaces fails after port flap Product-Group=junos	MACsec configured on logical interfaces (IFL) would fail to restore after a flap of the parent physical interface. This will lead to Cyclic Redundancy Check (CRC) errors and failure to pass traffic.
PR Number	Synopsis	Category: SRX4100/SRX4200 platform software
1626562	A major alarm DPDK (data plane development kit) Tx stuck issue of SRX4100/4200 Product-Group=junosvae	On the SRX4100 and SRX4200 platforms, it can detect DPDK (data plane development kit) Tx stuck issue and trigger a major chassis alarm goes which might trigger RG1 failover to the healthy node. A DPDK reset will be triggered only to the stuck port and if the reset resolves the tx stuck issue, the major chassis alarm will go off.
PR Number	Synopsis	Category: idp flow creation, deletion, notification, session mgr intfce
1705491	Network outage caused during change in IDP policy Product-Group=junos	Junos SRX platforms experience traffic outage during change in IDP (Intrusion Detection and Prevention) policy.
PR Number	Synopsis	Category: Interface Information Display
1636668	Management interface speed is incorrectly reported as 10G instead of 1G Product-Group=junos	On VM Host platforms, the management interface speed may incorrectly be displayed as 10G instead of 1G - this is a cosmetic issue.
PR Number	Synopsis	Category: Kernel software for AE/AS/Container
1649958	The user-defined speed does not take effect on the AE interface in certain scenarios on Junos platforms Product-Group=junos	On Junos platforms supporting GRES (Graceful Routing Engine Switchover), the bandwidth flag on the backup RE (Routing Engine) for the AE interfaces is set unconditionally, hence the bandwidth is struck at a value despite not being set by the user statically. When GRES is performed followed by link delete/add for the AE interface, the interface would get stuck with the bandwidth and gets synced with the backup RE. There can be seen partial service impact in the case of the RSVP (Resource Reservation Protocol) and the restoration to recover from this issue is to configure bandwidth manually and then remove the configuration.
PR Number	Synopsis	Category: Integrated Routing & Bridging (IRB) module
1685406	The protocol MTU for the IRB interface is not rolled back when the MTU of the IRB or IFD interfaces is modified or deleted Product-Group=junos	On all Junos platforms, the maximum transmission unit (MTU) on the Integrated Routing and Bridging (IRB) interface is not getting reset when the MTU configuration on IRB or IFD is removed. When MTU configuration is removed from the IRB interface, the internal data structure for the MTU configuration is not getting reset which might affect traffic and it is a rare case.
PR Number	Synopsis	Category: ISIS routing protocol
1696598	Wrong SRTE Secondary path weight makes the secondary path active in forwarding table Product-Group=junos	On all Junos and Junos Evolved platforms, SRTE (Segment Routing Traffic Engineering) Secondary LSP (Label Switched Path) should be only on standby in the forwarding table however it is also active and forwarding traffic due to the wrong metric calculation.
PR Number	Synopsis	Category: jdhcpd daemon
1693124	phone-home and SZTP may fail if phone-home daemon restarts Product-Group=junos	Onboarding device with Phone-home or SZTP fails in rare cases where phone-home daemon restarts.
1702831	DHCPv6 client options missing in solicit message if they exceed a certain length Product-Group=junos	On SRX platforms, when DHCPv6 options are configured, they are missing in solicit message if they exceed a certain length
PR Number	Synopsis	Category: Flow Module
1683835	The non-fragmented packets will get dropped on the SRX5K platforms with SPC3 card Product-Group=junos	On SRX5K platforms with SPC3 card and IFD enabled, when the fragmented and non-fragmented traffic are processed in a burst on the IPsec tunnel, the non-fragmented packets will get dropped. There will be a very small number of packet drops seen.
1699578	Application traffic drop seen on all SRX platforms due to TCP window size issue Product-Group=junos	On all SRX platforms, on rare occasions, after the SYN-ACK is established with the application server, the server will cause the TCP window size drop to zero after declaring a larger initial TCP window in the ACK packet. Though this is not prohibited but generally not expected behavior. Due to this, the packet will be dropped for that particular session.
1705996	A flowd process crash is seen on SRX4100/4200/4600, vSRX, and SRX5K with SPC3 card when a route is changed frequently Product-Group=junos	In a race condition, when PMI (power-mode-ipsec) is enabled and routes are changing frequently (3~5sec) on SRX4100/4200/4600, vSRX, and SRX5K with SPC3 card, a flowd process crash will be seen.
1708646	TCP session timeout seen on GRE tunnel Product-Group=junos	On GRE (Generic Routing Encapsulation) tunnel, TCP (Transmission Control Protocol) session lasts only 20sec and the session timeout seen due to high latency observed.
1708876	The IPv6 source-level fragmented SCTP packets passing through an IPSec tunnel will be dropped Product-Group=junos	On all SRX platforms, the IPv6 source-level fragmented Stream Control Transmission Protocol (SCTP) packets will be dropped when pass through an IPSec (Internet Protocol Security) tunnel.
PR Number	Synopsis	Category: SRX Firewall Authentication
1685116	The user authentication page is not rendering on the client browser Product-Group=junos	On all Junos SRX platforms due to cross-origin restriction access, the domain page does not load. The issue happened due to a firewall authentication failure.
PR Number	Synopsis	Category: High Availability/NSRP/VRRP
1678772	New secondary node to go into a disabled state after ISSU and failover RG0 because of fabric link failure Product-Group=junos	On SRX5K platforms with SPC3 card, the primary node will lose the backup node and fabric link failure is observed resulting in the new secondary node going into a disabled state after ISSU (In-Service Software Upgrade) and failover RG0 (redundancy group) in a chassis cluster environment.
1702763	The secure tunnel interface does not work properly in SRX standalone mode Product-Group=junos	On Junos SRX5400/5600/5800/4100/4200/4600/1500/vSRX3.0 platforms, when the secure tunnel interface (st0) st0.16000-st0.16385 is defined in standalone mode, st0.16000-st0.16385 does not work properly which leads to a traffic impact. From 20.4R1 onwards, st0.16000-st0.16385 is hardcoded to be added to a high-availability zone, so it will not work in other zones.
PR Number	Synopsis	Category: interfaces and zones for junos js software
1711729	The 'targeted-broadcast' feature will not work on some SRX platforms. Product-Group=junos	On SRX 1500, SRX4100, SRX4200 and SRX4600 based platforms running Junos, 'targeted-broadcast' feature will not work. As a result, features like wake-on LAN (WOL) which rely on targeted broadcast will be affected.
PR Number	Synopsis	Category: all logging related bugs on srx platforms
1631149	SRX5600/5800 - SNMP mib queries may result in occasional response timeouts Product-Group=junos	SRX5600/5800 - SNMP MIB queries for jnxJsSPUMonitoringMIB objects (1.3.6.1.4.1.2636.3.39.1.12.1.x) may result in occasional response timeouts
1708777	Setting the security log profile without a category or stream will lead to srxpfe process crash Product-Group=junos	On all Junos SRX platforms, when security log profile is added without a category or stream can cause the srxpfe to crash. Due to this there will be complete traffic loss.
PR Number	Synopsis	Category: Firewall Policy
1698508	Security policies go out of sync during ISSU Product-Group=junos	On all Junos platforms, policies go out of sync between RE (Routing Engine) and PFE (Packet Forwarding Engine) after performing ISSU (In-service software upgrade) leading to the wrong policy lookup.
PR Number	Synopsis	Category: IPSEC/IKE VPN
1696102	IPsec VPNs will disconnect after ISSU Product-Group=junos	On SRX platforms except SRX5K, & vSRX2.0/vSRX3.0, traffic will not pass over IPsec (Internet Protocol Security) VPN tunnels after ISSU (In-service software upgrade).
1699691	Mismatch in configured and negotiated proxy-identity parameters can lead to KMD core. Product-Group=junos	Mismatch in configured and negotiated proxy-identity parameters can lead to KMD core.
1704670	From 20.4 onwards, St0.16000 to st0.16385 will not be allowed to be configured in HA and MNHA mode Product-Group=junos	On SRX5000/SRX4100/4200/SRX4600/SRX1500/vSRX3.0 from 20.4 onwards, st0.16000 to st0.16385 will be disallowed to be configured in cluster HA (High Availability) and MNHA (MultiNode High Availability)mode. This range of st0 subunits will not be used for regular IPsec VPN traffic secure-tunnel interfaces and reserved only for HA link encryption internal usage.
PR Number	Synopsis	Category: Security platform jweb support
1681549	All the security policies on Junos SRX platforms can get deleted while trying to delete any particular policy via J-Web Product-Group=junos	On Junos SRX platforms, all the security policies under the 'security' hierarchy can get deleted while trying to delete a specific policy by applying a search filter via the J-Web interface. This issue is seen in Junos release 20.3R1 and above and results in traffic loss due to the deletion of policies.
PR Number	Synopsis	Category: Layer2 forwarding on EX/NTF/PTX/QFX
1716881	"set forwarding-options evpn-vxlan shared-tunnels" command will not be available for EX92XX and MX series platforms Product-Group=junos	Presence of "set forwarding-options evpn-vxlan shared-tunnels" configuration is causing mpc cores in EX92XX and MX series platforms
PR Number	Synopsis	Category: lacp protocol
1686546	The ethernet switching tables are not in sync between two PEs Product-Group=junos	MAC entries are not synced between the two PE (Provider Edge) devices as seen from the output of the 'show ethernet-switching table'. This is a rare timing issue which happens after GRES (Graceful Routing Engine Switchover) where the AE (Aggregate Ethernet) interface flaps leading to traffic loss.
PR Number	Synopsis	Category: Label Distribution Protocol
1666506	LDP session stuck in non-existent state when router has multiple addresses configured on loopback interface Product-Group=junos	When statically configured LDP neighbour and sessions transport address exists between two routers that is having multiple loopback addresses may cause LDP stuck in non-existent state. Deactivating and activating the LDP neighbour configuration can cause LDP session to remain stuck in the non-existent state.
PR Number	Synopsis	Category: Issues related to Junos licensing infrastructure
1686654	Subscribers are not able to connect to the device after the device reboot Product-Group=junos	Due to a rare timing issue all subscribers may fail to connect and get suck in init state after reboot of MX broadband network gateway (BNG).
1694935	license-check warning reported on backup RE by commit or commit check Product-Group=junos	On Junos Evolved platforms, license-check warning is reported on backup RE by "commit" or "commit check" operation, because license-check is not running in backup RE.
1710303	The installed software licenses tied to FPC1 were not visible after the NSSU upgrade Product-Group=junos	On Junos EX platforms with Virtual-chassis, when Nonstop Software Upgrade (NSSU) is being performed from master RE FPC1, after successful NSSU FPC0 becomes new master RE and software licenses tied to previous master RE FPC1 might get lost.
PR Number	Synopsis	Category: Port-based link layer security services and protocols that a
1711422	A dot1xd crash is seen on Junos EX2300 platforms Product-Group=junos	On Junos EX2300, if the Master Session Key (MSK) parameters are received from RADIUS (Remote Authentication Dial-In User Service) as part of authentication, a dot1xd crash could be observed which will impact the authentication on the dot1x enabled ports.
PR Number	Synopsis	Category: Application specific PRs (cos/snmp/time-sync/routing/BRAS)
1712790	The traffic is dropped while passing through VCP link on MX Virtual Chassis with MPC10 line card Product-Group=junos	The packets are being dropped when they come from the VC (Virtual Chassis) link into member in the MX Virtual Chassis with MPC10 line card.
PR Number	Synopsis	Category: SW PRs for MPC10E Interfaces
1642584	MPC10E: Quick 100G link-flaps has still some race conditions which can cause MQSS stream drain failures and xqss_sched_flush_queue failures Product-Group=junos	On MPC10E cards upon many very quick link down and up events in msec range might not always able to drain all traffic in the queue. This causes lost of traffic going through the interface. Traffic volume and class-of-service configuration does influence the exposure. See also PR1638410.
PR Number	Synopsis	Category: MPC11 ULC interface software related issues.
1698135	XQSS_CMERROR_DSTAT_INT_REG_DROP0_QDEPTH_UNDRN alarm is seen on MX2K platforms with MPC11E line cards upon aggregate interface down/flap event Product-Group=junos	On MX2010 or MX2020 platforms with MPC11E line cards, if a member link of an aggregate interface goes down or flaps, it might trigger "XQSS_CMERROR_DSTAT_INT_REG_DROP0_QDEPTH_UNDRN" alarm and disable-pfe action is executed. The affected PFE (Packet Forwarding Engine) will be disabled and no traffic will pass through it.
1703374	Some of the interfaces are going down on rebooting the MPC11E line card Product-Group=junos	The reboot of the MPC11E line card on MX2010/MX2020 platforms makes some of the interfaces down. Hence the traffic carried over the down interfaces will be dropped. This is caused by racing conditions between multiple processes during the addition of the interfaces after the FPC/PIC restart.
PR Number	Synopsis	Category: Multiprotocol Label Switching
1694648	The rpd crash will be observed during the MPLS label block allocation Product-Group=junos	On all Junos and Junos Evolved platforms in the Multiprotocol Label Switching (MPLS) scenario, the rpd crash will happen in rare conditions during MPLS label block allocation.
1701800	Memory leak issue in TED Product-Group=junos	On all Junos and Junos Evolved platforms, a memory leak is observed in TED (Traffic Engineering Database) when the inet tables is not cleaned up after routing instance deactivation.
PR Number	Synopsis	Category: For multicast snooping on MX
1711153	A crash can be observed for 'mcsnoopd' process when the VLAN name for igmp-snooping has certain characters Product-Group=junos	On all Junos and Junos Evolved platforms, a crash can be observed for the 'mcsnoopd' process. This issue is seen when the devices support the VLAN (Virtual Local Area Network) style of igmp (internet group messaging protocol)-snooping configuration(set protocols igmp-snooping vlan ) and the VLAN name begins with the word all and has certain other characters. The workaround is not to enable the snooping for such VLANs.
PR Number	Synopsis	Category: Multicast for L3VPNs
1675099	The multicast receiver receives no traffic in an extranet scenario having an SPT tree already established Product-Group=junos	On Junos and Junos Evolved MX/PTX platforms, if the SPT (Shortest-Path Tree) tree is already established on PE (Provider Edge) it sends (s, g, rpt) prune which will lead to no traffic to the multicast receiver in an extranet scenario.
PR Number	Synopsis	Category: MX Timing software
1696957	In the rare scenario, huge PTP Time errors are introduced and propagated to the downstream devices after the chassis reboot Product-Group=junos	On MX240, MX480, MX960, MX2010, and MX2020 platforms, the remote PTP (Precision Time Protocol) clock will recover the PTP clock with Time errors after a chassis reboot on the device running as Boundary Clock.
PR Number	Synopsis	Category: Microkernel for neo mpc
1670137	Multibit ECC error causes the whole MX platform chassis to go down Product-Group=junos	Faulty FPC (Flexible PIC Concentrator) on the MX platform chassis exhibiting multibit ECC (Error Checking and Correction) error (L2 cache error) will trigger this issue. The whole chassis goes down until the faulty FPC is removed from the chassis.
PR Number	Synopsis	Category: Track Mt Rainier RE NIC issues in Linux
1695794	The RE mastership switchover will not be triggered when the internal master interface on VM Host is down Product-Group=junos	On VM Host platforms, the Flexible PIC Concentrator (FPC) will be disconnected and the Routing Engine (RE) mastership will not be triggered when the master internal interface(eth1/eth2) is down. Traffic loss will be seen as FPCs are disconnected.
PR Number	Synopsis	Category: FreeBSD Kernel Infrastructure
1642287	Kernel crash is seen on the device if the device is continuously power cycled Product-Group=junos	This issue is applicable to EX3400/EX2300/EX4100. There is a possibility of kernel crash when the system will be in the process of coming up after reboot (and observed only with multiple iterations of continuous reboot cycles). This is observed only during the init sequence of the management driver and impact is limited to increased system boot time.
1700629	The command "request system zeroize" might not work properly on EX4400. Product-Group=junos	When issuing the "request system zeroize" command, an error message might be seen.
PR Number	Synopsis	Category: Kernel MPLS / Tag / P2MP Infrastructure
1702220	LDP flaps will be observed having LT interface with VLAN and LDP running between the logical-system instance and global instance Product-Group=junos	On Junos platforms, when the LT (Logical Tunnel) interface with VLAN and LDP (Label Distribution Protocol) is configured between the logical-system instance and global instance LDP flaps are observed.
PR Number	Synopsis	Category: TCP/UDP transport layer
1685113	BGP session flap with error BGP_IO_ERROR_CLOSE_SESSION Product-Group=junos	On all Junos platforms, a random BGP (Border Gateway Protocol) session flaps will be observed immediately after committing certain configuration changes with the error "BGP_IO_ERROR_CLOSE_SESSION" along with a TCP connection reset. This issue is generic to any TCP (Transmission Control Protocol) connection that has MD5 (Message-Digest Algorithm) enabled on it, eg targeted LDP Following error is also reported: rpd: bgp_pp_recv: rejecting connection from a.b.c.d (Internal AS xyz), peer in state Established rpd: bgp_pp_recv:5159: NOTIFICATION sent to a.b.c.d+64150 (proto): code 6 (Cease) subcode 5 (Connection Rejected)
1694463	The vmcore crash observed in low memory conditions Product-Group=junos	On all Junos platforms, when there are low memory conditions vmcore crash is observed which impacts the traffic.
PR Number	Synopsis	Category: OSPF routing protocol
1682506	Traffic loss is seen when the router in helper mode deletes the route for the router undergoing graceful restart Product-Group=junos	On all Junos OS and Junos OS Evolved devices, traffic loss is seen when the router in helper mode to support an OSPF graceful restart deletes the route for the router undergoing graceful restart. The device deletes the route due to the topology change.
1705975	OSPF routes are not getting installed after the interface is flapped Product-Group=junos	On all Junos and Junos Evolved platforms, the OSPF (Open Shortest Path First) routes are not getting installed when the interface is flapped multiple times. A deadlock is created when the OSPF routers wait for an LSA (Link State Advertisement) with a P2P (Point-to-Point) link to be advertised from its neighbouring router to add it to its router LSA and at the same time the neighbouring router also waits for an LSA with a P2P link to be advertised to add it to its router LSA. Both the OSPF routers advertise their router LSA without a P2P link and as a result, routes are not getting installed.
PR Number	Synopsis	Category: Express Chip L3 software
1650217	IRP memory parity issue might result in traffic loss on PTX and QFX10k Junos platforms Product-Group=junos	On PTX Series routers and the QFX10000 line of switches running Junos OS, parity issues might be observed related to Ingress Result Processor (IRP) memory might result in traffic loss.
PR Number	Synopsis	Category: Issues related to PKI daemon
1669426	Gradually increasing jsd memory usage will lead to jsd process crash Product-Group=junos	On all PTX and QFX platforms, gradually increasing jsd (Juniper Extension Toolkit (JET) service process) memory usage will lead to a jsd crash. It will restart automatically.
PR Number	Synopsis	Category: QFX access control list
1679574	Firewall functions will not work as expected when egress firewall filter is configured Product-Group=junos	On QFX5K/EX4400/EX4650 platforms, Egress firewall filter will not work as expected. There is no impact to traffic, but firewall functions will not work.
1692993	The fxpc coredump is generated and an FPC restart results in traffic impact Product-Group=junos	On all Junos platforms, when Ethernet Ring Protection (ERPS) is configured and the CLI command 'request support information' is executed, a coredump is generated and the FPC restarts, affecting traffic.
1695820	On QFX5110-VC-VCF platforms, traffic impact is seen when the firewall filter with DSCP action is enabled Product-Group=junos	On QFX5100-VC-VCF (Virtual-Chassis Fabric) when the firewall filter with action DSCP (Differentiated Services Code Point) is configured and is applied to any interface, the commit check gives a warning on FPC2 (Flexible PIC Concentrators) if the master has member id 0 and traffic drop can be seen.
1710704	FPC is down on QFX5k after committing an IPv6 filter Product-Group=junos	On Junos QFX5k series platforms, traffic is getting dropped as FPC (Flexible PIC Concentrators) is not up. The issue is seen when a 'flexible-match-range' is configured in an IPv6 firewall filter term and that filter is applied to any IPv6 interface. FPC goes down after committing the filter.
1710776	The message "fpc0 list_get_head, list has bad magic (0x0)" maybe output after the commit operation is complete Product-Group=junos	When a VACL filter is added or deleted, the message "fpc0 list_get_head, list has bad magic (0x0)" maybe output.
PR Number	Synopsis	Category: QFX L2 PFE
1694076	PFE crash is seen on all Junos QFX5K and EX46xx platforms with L2PT configuration Product-Group=junos	PFE (Packet Forwarding Engine) crashes on all Junos QFX5K and EX46xx platforms when L2PT (Layer 2 Protocol Tunneling) is configured on the interface having flexible-vlan-tagging with encapsulation extended-vlan-bridge. It causes a traffic impact.
1697835	PE device changes an outer tag-id in a local return environment Product-Group=junos	On Junos QFX5K platforms, PE (Provider Edge) device erroneously modifies an outer VLAN tag-id in local CE-to-CE communication. As the result, CE (Customer Edge) communication across local PE is not possible.
1706515	DCPFE crashes which leads to FPC restart Product-Group=junos	On Junos QFX and EX platforms, the DCPFE (Dense Concentrator Packet Forwarding Engine) crashed continuously after configuring the unsupported telemetry sensor '/junos/system/linecard/npu/utilization/' leading to the restarting of FPC. Once FPC is restarted, traffic is impacted.
PR Number	Synopsis	Category: QFX L3 data-plane/forwarding
1695847	Traffic forwarding fails when deleting all L2 related configurations Product-Group=junos	On QFX and EX series platforms, traffic loss will be observed for deleting all Layer2 related configurations when both L2 and Layer3 configurations are present.
PR Number	Synopsis	Category: QFX EVPN / VxLAN
1695058	Intra VLAN communication breaks in SP style config using VXLAN Product-Group=junos	Change from Enterprise (EP) to Service Provider (SP) style configuration result in reachability issue in pure L2 (Virtual Extensible LAN protocol) VXLAN setup.
1696979	VSTP will not work in the EVPN-VxLAN network Product-Group=junos	On all Junos QFX5K platforms, when using transparent mode EVPN-VxLAN (Ethernet Virtual Private Network-Virtual extensible LANs), the tagged BPDUs (Bridge Protocol Data Units) are dropped instead of being pushed into the VxLAN tunnel. This issue might affect all the services related to L2PT (Layer Two Protocol Tunneling).
1697244	Assigning VNI to VLAN will cause a small number of packets lost on other VLANs on the same interface Product-Group=junos	On Junos QFX5110 and QFX5120 devices, a small number of packets are lost when assigned VNI (VXLAN Network Identifier) to the VLAN (Virtual Local Area Network). There is a packet loss on incoming packets, but no packet loss on outgoing packets.
1707679	The spine does not reply to RS messages coming via the VXLAN tunnel in the CRB scenario Product-Group=junos	On Junos QFX5k/EX4650 platforms, when Router Solicitation (RS) packets come to the spine via the VXLAN tunnel in the centrally-routed bridging (CRB) scenario, the spine does not reply to the same and IPv6 Route Advertisement (RA) shows zero.
1712405	[EVPN-VXLAN] L3 VLANs created with IPv4 bits disabled Product-Group=junos	Traffic black-holing in EVPN (Ethernet VPN) fabric when performing configuration of new VLANs, it is found that L3 VLANs are having IPV4 bits disabled.
PR Number	Synopsis	Category: QFX5100 Interface related issues
1665800	Ports with SFP-T 1G plugged in may go to hung state on QFX5100 platforms Product-Group=junos	When the remote end server/system reboots, QFX5100 platform ports with SFP-T 1G inserted may go into a hung state and remain in that state even after the reboot is complete. This may affect traffic after the remote end system comes online and resumes traffic transmission.
PR Number	Synopsis	Category: QFX5100 Virtual Chassis
1671293	VC members are reloading randomly Product-Group=junos	On QFX5110 platforms, in a rare scenario, VC (Virtual Chassis) members are getting reloaded as the PRIVATE PFEMAN sockets related to the VC are getting dropped.
1694996	All members of the VCF will not reboot on QFX5k platforms Product-Group=junos	On QFX5k platforms with VCF(Virtual Chassis Fabric), when "request system reboot at now" CLI is executed from any member of VC then it will be pushed to all other members of VC. But sometimes member from which CLI is executed goes into a reboot so CLI can't be pushed to other members until it comes up again, when "request system reboot" without "at now" works as expected. This command should not be used in VC as some members will not reboot or there may be a time delay between reboot of members which will make VC unstable.
PR Number	Synopsis	Category: QFX5200/5110/5120/5210 Platform optics related issues
1660532	The port LEDs do not light up when 40G/100G physical interfaces are up. Product-Group=junosvae	The port LEDs do not light up when 40G/100G physical interfaces are up. This is a display issue. There is no service impact when this issue occurs.
PR Number	Synopsis	Category: QFX5200/5110/5120/5210 Platfom issues
1710855	The FPC will be offline after upgrading the system Product-Group=junosvae	On the Junos QFX5100 and QFX5120 platforms, after upgrading the system the PFE (Packet forwarding Engine) will crash continuously and Flexible PIC Concentrators (FPC) will remain down.
PR Number	Synopsis	Category: Category for QFX5K EVO Platform Software PRs related to Chas
1696113	The BFD session might be stuck in "Init" state on certain QFX5k platforms Product-Group=junos	On certain Junos QFX5k (QFX-5120-48Y/QFX-5120-32C/QFX-5120-48T) with multi-hop BFD (Bi-Directional Forwarding and Detection) enabled and multipath BGP (Border Gateway Protocol) is enabled, the BFD session can be stuck in "Init" state. This issue can be seen when there is a change in the routing path because of the interface flap for the BFD session. To recover from the issue, the BFD needs to be reconfigured.
PR Number	Synopsis	Category: RPD infrastructure issues related to NSR, GRES, switchover,
1701146	The rpd crash will be observed when any commit is performed Product-Group=junos	On all Junos and Evolved platforms with RSVP auto-mesh dynamic tunnel configuration, whenever a commit is done the rpd (Routing Process Daemon) crash will be seen. This happens due to null pointer access in the memory.
PR Number	Synopsis	Category: KRT Queue issues within RPD
1674310	The rpd crash may be encountered on Junos Evolved platforms Product-Group=junos	On Junos Evolved dual RE (Routing Engine) platforms, the rpd (routing protocol daemon) may crash if the NSR (Non-Stop Routing) switchover is performed before even the backup rpd is NSR ready. If the issue is encountered, the rpd may get stuck and not install routes. The workaround to avoid this is to perform the switchover when backup RE is fully synced.
PR Number	Synopsis	Category: Issue related to mcnh routing infrastructure within RPD
1658425	The next-hop does not be updated and errors observed when aggregated interface goes down Product-Group=junos	On all Junos OS Evolved platforms, dependency errors after interface flap are observed due to uncleared multicast composite next hop from RPD(routing protocol process daemon).
1708299	The rpd process crash is seen with scaled multicast next-hops Product-Group=junos	On all Junos OS Evolved platforms, due to timing issues in scaled setup with multicast next-hops, the rpd process might crash when the corresponding multicast next-hops Aggregated Ethernet (AE) interface flaps continuously. This will result in traffic loss.
PR Number	Synopsis	Category: RPD Next-hop issues including indirect, CNH, and MCNH
1652387	The rpd process might crash with back-to-back rpd restarts Product-Group=junos	On all Junos and Junos Evolved platforms, the rpd process might crash if the rpd process restarts back-to-back before the rpd process comes up.
PR Number	Synopsis	Category: Shard routing infrastructure within RPD
1699557	The rpd crash is observed when rib-sharding configured Product-Group=junos	On all Junos and Junos OS Evolved platforms, rpd crash observed while configuring BGP rib-sharding as route re-resolve happens due to another best match route. Traffic drop can be seen during rpd process recovery.
1710829	AIGP not distinguished with BGP-LU when rib-sharding is enabled Product-Group=junos	On all Junos platforms, BGP LU (Border Gateway Protocol Labeled Unicast) routes with AIGP (Accumulated Interior Gateway Protocol) metrics are not used when sharding is enabled, causing the incorrect route (having a high metric) to be selected as active.
PR Number	Synopsis	Category: Resource Reservation Protocol
1621771	RSVP refcount leak and the rpd crash observed post LSP churn Product-Group=junos	On all Junos and Junos Evolved platforms, rpd crash was observed due to RSVP refcount leak.
1661526	The rpd crash would be observed in a RSVP scenario Product-Group=junos	On all Junos and Junos Evolved platforms with RSVP enabled, the rpd (routing protocol daemon) crash will be observed while updating the RSVP path information with a new metric. When this happens, the core files are generated and some traffic loss will be seen.
PR Number	Synopsis	Category: RPM and TWAMP
1703104	The TWAMP test session packets are dropped when the payload is less than 52 bytes Product-Group=junos	On all Junos OS Evolved platform and Junos platform is using MPC10E/MPC11E/JNP10K-LC9600 linecards, If a Two-Way Active Management Protocol (TWAMP) client from a different vendor is transmitting a packet that is smaller than the 52 bytes then (the minimum allowed by RFC5357 is 41), the TWAMP reflector will be unable to function. Any packets from the TWAMP client will therefore be ignored. Similar to this, juniper boxes will also block any answers that are less than 52 bytes long, causing TWAMP test sessions to report a packet loss.
PR Number	Synopsis	Category: Generic platform and infra issues for MS-MIC and MS-MPC(XLP)
1699115	Transit tunnels fails and remains down on all Junos based MX and SRX platform with IKE-NAT-ALG enabled Product-Group=junos	On all Junos based MX and SRX platforms, the transit tunnels are not getting established and remained down due to a 10240 IKE (Internet Key Exchange) cookie limitation on MX-SPC3 and 10240 cookie limitation on the SRX platform. IKE tunnel sessions are getting dropped on the device and caused a traffic impact.
PR Number	Synopsis	Category: SFW, CGNAT on MS-MIC/MS-MPC (XLP)
1701305	On Junos platforms with MS-MPC cards the IKE ALG inactivity timeout value stays fixed Product-Group=junos	On Junos platforms having MS-MPC (Multi-Service Modular PIC Concentrator), the value for the inactivity timeout for IKE (Internet Key Exchange) ALG (Application Level gateway) is automatically set to 14400 seconds even though a custom timeout value is configured by the customer. This issue is seen only in the case of IKEv1 having SFW (Stateful Firewall) without any NAT (Network Address Translation) configuration. When this issue is seen IKE transit tunnel sessions remain in the session table longer than expected, not honoring the configured timeout value. This issue is self-recoverable.
PR Number	Synopsis	Category: SRX branch platforms
1713759	Continuous vmcores observed on the secondary node when committing the "set system management-instance" command Product-Group=junos	On Junos SRX3xx series platforms, when the "set system management-instance" command is committed on the secondary node, continuous vmcores are observed on primary and secondary nodes. No recovery action is needed for the primary node and the secondary node of the cluster reboots automatically to recover from the error. The cluster redundancy is not restored until the management-instance knob is removed using the "delete system management-instance".
PR Number	Synopsis	Category: MX10003/MX204 Linux issues (including driver issues)
1660446	MX204 - SSH non-default port configuration causes FPC offline after upgrade to 21.4 Product-Group=junos	For MX204, MX10003, ACX5448 platform, if a non-default ssh port is configured for system login, after upgrade to 21.4 release, the FPC is stuck in offline. To avoid such issue please use default SSH port and use protect RE filter to only allow the access from the trusted source.
PR Number	Synopsis	Category: MX10003/MX204 MPC defects tracking
1701183	Traffic loss is seen due to interface flap when changing speed from 10G and 1G Product-Group=junos	On Junos MX204/MX10003, traffic loss is seen on 4x10G Channelized interfaces when the speed is changed from 10G to 1G on one lane which causes the other 3 lanes to flap and stay down.
PR Number	Synopsis	Category: SRX-1RU platfom chassisd SW defects
1711467	SRX4600 doesn't support ae interfaces Product-Group=junos	On Junos SRX4600 platforms, ae interfaces doesn't come up or not work properly when configured in HA (High Availability) cluster mode.
PR Number	Synopsis	Category: SRX-1RU platfom related protocol, QoS, filtering features et
1683213	SRX4600HA might not failover properly due to a hardware failure Product-Group=junosvae	In SRX4600 platform, there is a possibility of raising alarm function might not work in FPGA failure. It might affect device availability.
1712727	Continuous vmcores observed on the secondary node when committing set system management-instance command Product-Group=junos	On Junos SRX4600 device, when the "set system management-instance" command is synced to the secondary node, continuous VMcores are observed on primary and secondary nodes. The secondary node of the cluster reboots automatically to recover from the error. The cluster redundancy is not restored until management-instance knob is removed using "delete system management-instance"
PR Number	Synopsis	Category: ZT/YT pfe firewall software
1692070	The firewall bridge filter policers (attached to AE interface) are not working on all Junos MX platform with MPC10 card upon deactivate-activate a term intended to limit overall traffic Product-Group=junos	Traffic policing will not work as expected on all Junos MX platform with MPC10 card after doing deactivation-activation of the term intended to limit overall traffic when the firewall and policers are applied to the bridge unit using shared-bandwidth-policers on AE (Aggregated Ethernet) interface.
1701320	Traffic loss is seen on MPC10E due to null pointer access without any safe check Product-Group=junos	On the Junos MX series platforms with MPC10/11, when the line card is inserted, it restarts and traffic loss is seen generating the core dump due to the null pointer access without any safe check.
PR Number	Synopsis	Category: ZT/YTpfe bridging, learning, stp, oam, irb software
1668837	EVPN PE router might respond traceroute with unexpected source IP address to remote CE Product-Group=junos	On all Junos MX platforms with MPC1/2/3/4/ 5/6/ 7/8/9 line cards. IP addresses of the transit nodes listed by traceroute might be incorrect in a certain scenario: when the path was formed with EVPN type-5 route.
PR Number	Synopsis	Category: Trio pfe qos software
1706494	Severity reclassification of queuing ASIC XQSS and memory parity error auto recovery Product-Group=junos	On Junos with MX platform using specific MPC7E/MPC8E/MPC9E/JNP10K-LC2101/JNP10003-LC2103/JNP10K-LC480 line cards in the event of a transient memory hardware issue, parity errors are generated which cannot be rectified by the ASIC. When the Flexible PIC Concentrators(FPC) encounters such an error, it will automatically deactivate the Packet Forwarding Engine(PFE) which leads to traffic impact.
PR Number	Synopsis	Category: Trio pfe stateless firewall software
1666966	The MPC hosting an AE member interface with a shared bandwidth policer configured at the AE could crash upon encountering an HMC fatal error Product-Group=junos	MX MPC7/8/9, MX204 MPC, MX10003 MPC, JNP10K-LC2101, EX9200 MPC hosting an Aggregated Ethernet (AE) member interface with a shared bandwidth policer configured at the AE could crash upon encountering a Hybrid Memory Cube (HMC) fatal error.
PR Number	Synopsis	Category: Trio pfe l3 forwarding issues
1655651	Telemetry is reporting In-Errors when knob "ignore-l3-incompletes" is configured Product-Group=junos	On all Junos MX platforms, Telemetry reports In-Error when "ignore-l3-incompletes" knob is configured.
PR Number	Synopsis	Category: Issues related to port-mirroring functionality on JUNOS
1683192	Traffic loss is seen with port-mirroring is enabled on AE interface in multicast downstream Product-Group=junos	On MX and EX Platforms with Trio based MPCs, traffic loss is seen as multicast traffic will not be replicated to all downstream ports when port-mirroring is enabled on the AE (Aggregated Ethernet) interface with the destination port as another AE.
PR Number	Synopsis	Category: UI Infrastructure - mgd, DAX API, DDL/ODL
1679413	Configuration filtering doesn't work when the logical system is present Product-Group=junos	On SRX platforms, configuration filtering doesn't work when the logical system is present.
1699245	The mgd process might crash during commit synchronize Product-Group=junos	On all Junos and Junos Evolved platforms with dual Routing Engines, during commit script synchronize, the backup Routing Engine mgd get process crashes once the memory leak reaches 65K. When this happens, commit configurations from the primary Routing Engine fail.
1708321	MX960 :: CST:RE goes to amnesiac state, when rebooting the DUT -mgd: error: translation script failure Product-Group=junos	When a toggle attribute with variable is committed under dynamic-profiles hierarchy, upon reboot system will go to amnesiac mode
PR Number	Synopsis	Category: Issues related to Logging/Tracing, errmsg, eventd infrastruc
1689661	source-address on syslog at custom routing-instance not applied right after rebooting Product-Group=junos	syslog may not be sent out via configured source address when target host exists on a custom routing-instance
PR Number	Synopsis	Category: Junos Fusion Satellite Device Infrastructure
1682680	The Junos Fusion Satellite device will be stuck in the SyncWait state Product-Group=junos	On Junos MX, EX, and QFX platforms, post rebooting the device the Junos Fusion Satellite devices are trying to generate the ssh keys before clearing the old keys. The Satellite devices will not be responsive as they are stuck in the SyncWait state leading to traffic loss.
PR Number	Synopsis	Category: PTX/QFX100002/8/16 interface software
1712007	The interface does not come up or keeps flapping Product-Group=junos	On Junos PTX10008/PTX100016 devices with LC1101/LC1102 line cards, any event that causes an interface state change can lead to interface flapping or the interface may not come back at all. This leads to traffic impact on that interface.
PR Number	Synopsis	Category: For GPRS security features on highend SRX series
1704472	GTPv2 Message Filtering is not working Product-Group=junos	GTPv2 Message Filtering not working - resolved
PR Number	Synopsis	Category: Virtual Router Redundancy Protocol
1680178	VRRP master-master condition might occur when there are more than two devices in the VRRP group Product-Group=junos	When more than two devices are part of the VRRP group and a lower priority master sends a keepalive as master, low priority backup may transition to master state and get stuck.
1697394	VRRP Master session on AE ifl having child links on Satellite Device stops transmission post GRES Product-Group=junos	On Junos Fusion Aggregation Device platforms, when the VRRP (Virtual Router Redundancy Protocol) interface is configured on IRB (Integrated Routing and Bridging) with underlying L2 interface on Satellite Device post GRES (graceful Routing Engine switchover) VRRP Master stops transmission of VRRP PDU (Protocol Data Unit) resulting in VRRP Master/Master state.
PR Number	Synopsis	Category: Xellent Platform issues
1709817	Ports with QSA adapter are down Product-Group=junos	On Junos PTX1000 and PTX10002-60C/QFX10002-60C platforms, ports which use the QSA (QSFP-to-SFP Adapter) may not come up when running software version containing the fix for PR 1620527.

21.4R3-S3 - List of Known issues

See TSB70539 [juniper.net]

Modification History

2023-03-17 First publication