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.
Infrastructure
-
NTP time drift on the affected Junos releases. Earlier implementation of kvmclock with vDSO (virtual Dynamic Shared Object) which helps avoid the system call overhead for user space applications had problem of time drift, the latest set of changes takes care of initializing the clock after all auxiliary processors are launched so that the clock initialization is accurate. PR1691036
Interfaces and Chassis
-
Following two Failure messages seen brcm_rt_ip_mc_ipmc_install:2455 Failed (Invalid parameter:-4) This message is due to IPMC Group being used is not created, when RE tried to add this check indicates there is a parameter mis-match. brcm_rt_ip_mc_ipmc_install:2455 Failed (Internal error:-1) This message is due to Failure to read IPMC Table or any memory/register. PR1461339
-
On all Junos platforms, if a speed mismatch happens in the LAG (Link Aggregation) and member interface then a traffic drop will be seen. PR1725168
Layer 2 Features
-
In case of the access-side interfaces 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 20 to 50 minutes traffic drop on the existing logical interface. PR1367488
Layer 2 Ethernet Services
-
On QFX5100 and QFX5110 devices, vendor-id format maybe incorrect for network ports. This does not impact the ZTP functionality or service. The DHCP client config is coming from two places, i.e AIU script and vsdk sandbox. The DHCP client config coming from AIU script has the serial Id in vendor id where as the default config from sandbox does not have. PR1601504
Platform and Infrastructure
-
On the QFX5100 line of switches, inserting or removing optics on a port might cause a Packet Forwarding Engine Manager CPU spike and an eventual microcode failure. PR1372041
-
VXLAN VNI (multicast learning) scaling on QFX5110 traffic issue is seen from VXLAN tunnel to Layer 2 interface. PR1462548
-
In the platform using INH (indirect next hop, such as Unilist) as route next hop type for multiple paths scenario (such as BGP PIC or ECMP), the session fast-reroute might be enabled in Packet Forwarding Engines (PFEs). When the version-id of session-id of INH is above 256, the PFE might not respond to session update, which might cause the session-id permanently to be stuck with the weight of 65535 in PFE. It might lead PFE to have a different view of Unilist against load-balance selectors. Then either the BGP PIC or the ECMP-FRR might not work properly and traffic might be dropped or silently discarded. PR1501817
-
When launching a guest Virtual Machine (VM) to run a third party application on Junos OS 15.1R1 and above, the guest VM might be shown as "UNAVAILABLE" even after successfully installing the third party application. This is due to duplicated device ID assigned to different disks. PR1529596
-
On QFX5110 VC, FPC may gets disconnected with 24K DHCPv6 relay scaling, after the traffic is stopped. The
pfe_listener_disconnect
error messages might get generated. PR1594748 -
Pim VXLAN does not work on the TD3 chipsets enabling VXLAN flexflow after the Junos OS Release 21.3R1. Customers Pim VXLAN or data plane VXLAN can use Junos OS Release 21.3R1. PR1597276
-
On all devices running Junos OS or Junos OS Evolved, where this is a high BGP scale with flapping route and the BGP Monitoring Protocol (BMP) collector/station is very slow, the rpd process might crash due to memory pressure. PR1635143
-
When MACSEC and VRRP are enabled on QFX5120 VC, MACSEC sessions are flapping at random times. Without VRRP this issue is not seen. PR1640031
-
On all QFX5100 Virtual Chassis platforms, after the reboot, Virtual Chassis port (VCP) ports may not establish a VCP connection and Cyclic Redundancy Check (CRC) errors are also observed. PR1646561
-
On QFX platform, v6 ifl stats are being derived from the underlying ifd stats unlike on PTX where they are hardware assisted. Hence, they are not very reliable and are at best, guesstimate. PR1653671
-
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. PR1665800
-
On QFX5100 platforms (both stand-alone and VC scenario) running Junos, occasionally during the normal operation of the device, PFE (Packet Forwarding Engine) can crash resulting in total loss of traffic. The PFE reboots itself following the crash. PR1679919
-
On Junos QFX5100 and EX4600-Virtual Chassis (VC) and Virtual Chassis Fabric (VCF) platforms on upgrading Virtual Chassis Fabric (VCF) and toggling the interface, when FPC (Flexible PIC Concentrators) is disabled and rebooted, the member fails to join the virtual chassis and the interface remains disabled even after been enabled. PR1689499
-
When TISSU upgrade is done from 22.4 release onwards, the box come up as backup RE. Work-around:- To make is master following command needs to be run again. sysctl -w hw.lc.issuboot=0 sleep 10 sysctl -w hw.re.issu_state=0 sleep 10 sysctl -w hw.re.tissu=0 sleep 10 sysctl -w hw.product.pvi.config.chasd.no_re_status_on_backup=1 sleep 60. PR1703229
-
The
show chassis hardware
command indicates duplicate entries for PSU and FAN tray after USB clean install or zeroize. PR1704106 -
On Junos and Junos OS Evolved platforms, the dcpfe(Dense Concentrator Packet Forwarding Engine) process crash will be observed due to memory fragmentation issue. This is a very rare case and would impact traffic as due to dcpfe failure the PFE restarts, so the interfaces flaps. PR1711860
-
On EX4650 and QFX5120-Y devices, the 10G interfaces are not coming up simultaneously when different Small Form-factor Pluggable(SFPs - 10G and 1G) are plugged in within the same 4 port group. Normally 10G interface by itself will be up when set to 1G if no other SFP is plugged in. PR1714833
-
In a VC of QFX5100-24Q with an expansion module EX4600-EM-8F, if VC is formed on 10G ports then after the reboot of VC, the 10G connections will be lost and the line card will show as not present. This will impact traffic on the 10G ports after connection is lost. PR1718062
-
On the Junos QFX5200 platform, sometimes 100G link will go down and will remain down. PR1725116
-
This problem is seen on QFX10000 platform, when ingress Sflow is enabled on bridged AE (tagged) interfaces and that incoming traffic gets forwarded via IRB interfaces on an ECMP path. In ECMP scenario, Sflow injects TAL request into the ASIC which are resulting in the traps tracked via this PR. Hence the forwarding traffic is unaffected. PR1729316
-
This problem is seen on QFX10k platform, when egress Sflow is enabled on bridged AE (tagged) interfaces. In this case traffic gets forwarded via IRB interfaces to bridged AE (tagged) interfaces and to ECMP path. In case of ECMP, Sflow injects TAL request into ASIC which are resulting in the traps tracked via this PR. Original forwarding traffic is unaffected due to this problem. PR1730882
-
The QFX10000 devices forwards a packet to a remote Vtep in MAC VRF even though no matching VLAN ID is found on the source port. PR1735961
-
On all Junos based QFX platforms, when sflow is enabled with ECMP, ddos-protection violation of protocol group resolve is triggered. Sampled packets will be dropped and sflow will stop sending packets to the collector. This is a non-service impacting issue, however sflow will be impacted. PR1741461
Virtual Chassis
-
On Junos QFX5100 platforms running qfx-5e images in Virtual Chassis setup, when Virtual Chassis Port (VCP) links are connected between PHY and PHYLESS ports, CRC alignment errors will be seen. As a result, there can be traffic loss on these links. PR1692102