Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?


Open Issues

Learn about open issues in this release for QFX Series switches.

For the most complete and latest information about known Junos OS defects, use the Juniper Networks online Junos Problem Report Search application.


  • End-hosts might not communicate via EVPN-VXLAN domain after Ethernet Segment Identifier (ESI) failover. This issue affects QFX5000 platforms only. PR1584595

  • With shared-tunnels enabled, assisted replication (AR) tunnels are not formed on Data Center Gateway (DC-GW) for the MAC-VRF instances. PR1584790

  • In a multihomed and EVPN-VXLAN scenario, when MAC and MAC-IP are local and remote on two PE devices respectively, then proxy MAC generation will kick in. In such scenario, when the MAC timer expires, the MAC entry does not get cleaned up. PR1612866

Interfaces and Chassis

  • After restart of ICCP service, the show iccp command takes more than 20s to display the client-name information. PR1618987

Layer 2 Features

  • If the access-side interfaces are used as SP-style interfaces, when a new logical interface is added and if there is already a logical interface on the physical interface, there is a 20-50 ms traffic drop on the existing logical interface. PR1367488

Layer 2 Ethernet Services

  • It is observed rarely that issuing the request system zeroize command does not trigger ZTP. As a workaround, reinitiate the ZTP. PR1529246


  • The rpd process generates core files at rt_iflnh_set_nhid due to assertion caused by failure of hbt_insert for nhid belonging to a logical interface. It is seen that there is a duplicate entry present which causes the hbt_insert failure. PR1588128

Network Management and Monitoring

  • The dfwd core files are generated when accessing ephemeral date base files which is deleted through script. PR1609201

Platform and Infrastructure

  • The commit synchronize command fails because the kernel socket gets stuck. PR1027898

  • When a VLAN is added as an action for changing the VLAN in both ingress and egress filters, the filter is not installed. PR1362609

  • The unified ISSU is not supported on QFX5200 switches and fails from Junos OS Release 17.2X75-D43.2 to some target versions. Also, dcpfe crash might be seen. PR1438690

  • On QFX5000 platforms with the instance-import statement configured, deleting route which has the next-table configuration might result in an unexpected route next hop. PR1477603

  • The show pfe filter hw filter-name filter name command fails to retrieve the Packet Forwarding Engine programming details of the filter. PR1495712

  • When the DHCP relay mode is configured as no-snoop, the offer gets dropped due to incorrect ASIC programming. PR1530160

  • FIPS mode is not supported. PR1530951

  • On QFX5100 platforms not running the QFX-5E codes (non TVP architecture), when an image with third party SDK upgrade (6.5.X) is installed, the CPU utilization might go up by around 5 percent. PR1534234

  • The Socket to sflowd closed error messages come up when the ukern socket to sflowd daemon (server) is closed. The error is rectified automatically as the client successfully reestablishes the connection in the subsequent attempts. When these errors are consistent, it indicates a communication issue between sflowd and the sFlow running on the FPC. PR1538863

  • On an EVPN-VXLAN scenario, vmcore files are generated on the master and backup Routing Engine of the QFX10008 platforms with Layer 2 or Layer 3 multicast configuration. PR1539259

  • 100 G AOC from third-party does not come up after multiple reboots. It recovers after interface enables or disables. PR1548525

  • On QFX5100-48S platforms, the interface might remain in down state after loading the QFX 5E Series image on the device. This issue is only observed with 1G optics (SFP-SX and SFP-LX10) and when the auto-negotiation setting is enabled. The traffic through the affected interface will be lost. PR1554098

  • On QFX10000 line of switches except QFX10002 platform, when the Linux Chassis Manager (LCMD) polls power supply modules (PSMs), the memory used for that polling does not get freed. The amount of memory that is not being freed depends on the number of sensors (FPCs and PICs) installed in the chassis. The LCMD process will continue to consume memory until all of the free memory available to VM Host has been exhausted. At that point, the LCMD restarts and causes mastership switchover of the Routing Engine. PR1555386

  • 5M DAC connected between QFX10002-60C and MX2010 platforms does not link up. But with 1M and 3M DAC, interoperability works as expected. Also, it is to be noted that connection between QFX10002-60C and ACX or traffic generator works seamlessly with the same 5M DAC. PR1555955

  • To avoid the additional interface flap, interface hold time needs to be configured. PR1562857

  • Unable to execute python or shell scripts in flex mode. Starting in Junos OS Release 21.1R1 release, Junos OS ships with python3 (python2 is no longer supported). In ZTP process, if a python script is being downloaded, ensure that the python script follows python3 syntax (there are certain changes between python2 and python3 syntax). Also, so far (that is, until Junos OS Release 20.4R1), the python script had #!/usr/bin/python as the first line (that is, the path of the python interpreter). The same needs to be changed to #!/usr/bin/python3 from Junos OS Release 21.1R1. PR1565069

  • The chassisd logs are flooded with the pic_create_ifname: 0/0/0 pic type F050 not supported error messages for every port that is connected. This happens repeatedly in a few seconds. PR1566440

  • BUM traffic replication over VTEP is sending out more packets than expected and there seems to be a loop. PR1570689

  • On QFX10K2-60C platforms, disk missing alarms are not seen. PR1573139

  • The OSPF session over IRB might not come up in an EVPN-VXLAN scenario. PR1577183

  • On an EVPN-VXLAN scenario, after performing the run restart l2-learning, IS-IS sessions were not established. PR1580971

  • On QFX5100 platforms, while checking the DHCP smart relay over IRB interfaces, the renew-ack might not be seen in the DHCP client. PR1581025

  • When soft loopback port and analyzer configurations are committed together, hardware might not get programmed with the analyzer. This issue is not seen when a physical loopback is used to achieve the same. PR1581542

  • In a fully loaded device, firewall programming fails at times due to scaled prefix configuration with more than 64800 entries. PR1581767

  • When physical loopback is used and both the ports are with EP style in the same RSPAN VLAN, it might lead to flooding. PR1581876

  • On all Junos OS platforms, when there is a congestion on the link where telemetry streams are connected, then in a race condition, na-grpcd core files are generated. This impacts the telemetry service as the na-grpcd will take a minute to come back online. PR1587956

  • On QFX5000 line of switches, the Flexible PIC Concentrator (FPC) or dcpfe process might go into a very uncommon state when multiple third-party counter (bcmCNTR) threads are running or spawned in FPC. This state causes the dcpfe process to crash or the FPC to reboot. The purpose of bcmCNTR is to poll statistics from hardware. PR1588704

  • On QFX5100, QFX5110, QFX5120, QFX5130, QFX5200, QFX5210, and QFX5220 platform switches with third party chip as Packet Forwarding Engine, if IS-IS is enabled on an integrated routing and bridging (IRB) interface and the maximum transmission unit (MTU) size of the IRB interface is configured with a value great than 1496 bytes, the IS-IS hello (IIH) PDUs with jumbo frame size (that is, great than 1496 bytes) might be dropped and does not sent to the IS-IS neighbors. PR1595823

  • DCI InterVNI and IntraVNI traffic might silently dropped and discarded in GW node with egp.v4_chksum trapstats due to tagged underlay interfaces. To prevent this, do not use tagged underlay at gateway nodes in DCI based topology. PR1596462

  • Post unified ISSU, issue will be seen on deleting and adding back sample configuration below: interfaces { ge-0/2/5 { flexible-vlan-tagging; encapsulation flexible-ethernet-services; gigether-options { ethernet-switch-profile { tag-protocol-id 0x0800; } } unit 1 { vlan-id 1; family inet { filter { input inet4filter; output inet4filter; } address; } family inet6 { filter { input inet6filter; output inet6filter; } address 2002:0000:0000:0000:0000:0000:1000:0641/126; } } } }. PR1596483

  • During FRR, when more than one multihome interface is down, traffic might get loop for QFX5110 platforms. PR1596589

  • On QFX Series platforms, the dcpfe process or FPC might crash during boot time if you reboot the devices. PR1597479

  • Read write lock is not acquired during the sysctl invocation. The assert triggered in the interface state function call leads to go Routing Engine 1 to debug (db>) prompt. PR1598814

  • On QFX5200 platforms, dcpfe core files are generated while testing ISSU. PR1600807

  • The convergence time degradation is seen in IS-ISv6, OSPFv2, and OSPFv3 when comparing convergence time with Junos OS Release 21.1R1.5. As it is a convergence time issue, many components are involved and hence need investigation of rpd, kernel, and Packet Forwarding Engine. PR1602334

  • On QFX5120 switches, traffic loss might be seen when primary link is disabled with aggregated Ethernet link protection configuration. PR1604350

  • On QFX10008 platforms. the system reboot takes approximately 9 minutes for FPCs to come online after system reboot command is issued. PR1605002

  • On the QFX5110 and QFX5120 platforms, when you configure an EP and SP style logical interface on same physical interface with native VLAN on SP style logical interface in an EVPN-VXLAN scenario, traffic drop or unexpected behavior in traffic might be seen. PR1606106

  • Removing and adding Virtual Chassis ports might cause any of the Virtual Chassis ports to remain down. PR1606705

  • On QFX10000 line of switches, service provider style configuration is not supported with sFlow. PR1608360

  • On QFX10002-60C switches under MAC statistics, the output-mac-control-frames and the output-mac-pause-frames do not increment. PR1610745

  • On QFX5100 Virtual Chassis, when 118 (max) lag groups are configured. then there might be traffic loss of a few packets intermittently. PR1611162

  • On QFX10002-60C platforms in a scaled EVPN-VXLAN environment (for example, more than 3000 VLANs without assisted replication (AR) or more than 2000 VLANs with AR), continuous FPC crash and dcpfe core might be observed. PR1612871

  • On QFX5120-48Y switches, when scaled and baseline configurations are loaded multiple times one after other without much waiting time in between, then the traffic or protocols on pure Layer 3 interfaces might behave in an undefined or unexpected manner. PR1612973

  • On QFX5000 Series platforms in an EVPN-VXLAN scenario, when remote VTEPs that are part of multihoming ESIs go down, the ARP entries that depend on these VTEPs are not reprogrammed correctly under certain circumstances. The next hop resolutions might be left incomplete. The BFD session get stuck in init state after l2-learning restart due to incomplete ARP resultions. This causes the traffic loss. PR1618280

  • PIM VXLAN is not working on TD3 chip sets enabled with VXLAN flexflow after Junos OS Release 21.3R1. Customers using PIM VXLAN or data plane VXLAN can use the Junos OS Release 21.3R1. PR1618829

  • The NSSU option is not available from Junos OS Releases 21.2R1. This option is missing from the time the UI component publish has been separated out. PR1621611

Routing Protocols

  • In a Virtual Chassis or Virtual Chassis fabric scenario, inconsistent MCSNOOPD core file is seen when the igmp-snooping configuration is removed. PR1569436

  • The issue happens in a PRPD based setup where BMP is configured. Upon clearing or deleting a peer, peer cleanup process gets stuck and peer stays in idle (close in progress) state forever. PR1618103