Alert Type

SRN - Software Release Notification

Product Affected

PTX QFX

Alert Description

Junos Software Service Release version 20.3R1-S1-EVO 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

Solution

Junos Software service Release version 20.3R1-S1-EVO is now available.

20.3R1-S1-EVO - List of Fixed issues
PR Number Synopsis Category: Express BT PFE L3 Features
1539864 On channelizing 40G to 4x10G, ping is failing on the channels.
Product-Group=evo 		In PTX10001-36MR when ports are used with 40G optics in 4x10G channelized mode, then ping fails on channel2 and cahnnel3 for following ports - et-0/1/9, et-0/1/10, et-0/1/11, et-0/2/1, et-0/2/2, et-0/2/3 ,
PR Number Synopsis Category: System Management High availability
1539705 On re0 cli node reboot, rpd is unable to connect to snmpd and hence we see scheduler slips in rpd and protocol sessions flap
Product-Group=evo 		On re0 cli node reboot, rpd is unable to connect to snmpd and hence we see scheduler slips in rpd and protocol sessions flap
PR Number Synopsis Category: Interface PRs defect & enhancement requests
1540431 FTI tunnel commit fails for MTU change on its family(inet, inet6)
Product-Group=evo 		On Junos EVO platforms, FTI(Flexible Tunnel Interfaces) tunnel MTU change fails with family inet/inet6 on PTX10001/10008/PTX10016.
PR Number Synopsis Category: Express PFE CoS Features
1536963 Triton multicast driver is incomplete
Product-Group=evo 		Triton multicast software is incomplete up to 20.3R1-EVO and does not work correctly for the BT.dp1 half of ports per BTchip. See LC1201 and PTX10001-36MR hardware product diagrams for affected ports.
PR Number Synopsis Category: Express PFE L3 Multicast
1535545 Scapa : Multicast - After FPC restart observing both traffic flood and loss towads downstream ae receiver interfaces
Product-Group=evo 		When multicast is over AE bundle which spans multiple FPCs, if one FPC is restarted, the multicast traffic can be duplicated or loss on the receivers.
1537700 On a PTX10008 running Junos Evolved, running PIM Dense Mode (DM) may cause an FPC to restart unexpectedly.
Product-Group=evo 		On a PTX10008 running Junos Evolved, running PIM Dense Mode (DM) may cause an FPC to restart unexpectedly.
1538849 Nexthop "addCollision" Errors are seen with scaled multicast routes
Product-Group=evo 		During a multicast routes convergence even, you may see "Nexthop addCollision" error messages.
1539152 Auto-RP not getting learnt when a PTX10008 node is in discovery mode
Product-Group=evo 		In some network topology, Auto-RP might not be getting learnt when a PTX10008 node is in discovery mode.
1539912 Scapa : Multicast traffic null route on downstream adjacent node seen for ~5-6 sec as dlu.ucode.inv_start_pc trap after AE member link delete / add on Scapa node
Product-Group=evo 		When new member is added to the AE bundle which has other members on the same fpc, traffic loss can happen if multicast traffic is running over the ae bundle.
PR Number Synopsis Category: ISIS routing protocol
1482983 The output of the show isis interface detail command might be incorrect if wide-metrics-only is enabled for IS-IS and the ASCII representation of the metric in decimal is more than 6 characters long.
Product-Group=evo 		If 'wide-metrics-only' is enabled for any IS-IS level and a metric configured on the IS-IS enabled interface for that level has ASCII representation in decimal more than 6 characters long, this interface's metric for that level will be merged with 'priority' field value in the output of 'show isis interface detail'.
PR Number Synopsis Category: Issues related to Junos licensing infrastructure
1540881 Installed licenses are getting deleted after image upgrade
Product-Group=evo 		In earlier Junos Evolved OS releases, the license keys do not persist. When performing Junos Evolved OS upgrade or downgrade the customer must reinstall the license keys. From the 20.3R1-S1 release onwards, the persistence issue is resolved. Junos Evolved OS upgrade or downgrade does not require reinstalling the license keys.
PR Number Synopsis Category: Multicast Routing
1536593 rpd crash @ msdp_listen_start during upgrade with MSDP configuration loaded on DUT.
Product-Group=evo 		rpd crash @ msdp_listen_start during upgrade with MSDP configuration loaded on DUT.
PR Number Synopsis Category: RPD Next-hop issues including indirect, CNH, and MCNH
1535680 AE interfaces stuck in detached state after config delete/rollback with multicast config
Product-Group=evo 		After rollback to baseline config, it will bring down all interfaces including ae-ifls. The ifcomplist associated with this ae is removed too, however RPD is in middle of adding an comp-nh but it will send a wrong ifcomplist which is no more valid. Comp->Uncast->IFL(ae), till unicast is freed IFL will be struct in down state.
PR Number Synopsis Category: PTX10K platform specific fabric PRs
1540576 PTX1008: 'SIB 0 ZF0 internal temp' sensor normal fan speed temperature threshold is set to 0.
Product-Group=evo 		PTX1008 : 'SIB 0 ZF0 internal temp' sensor Normal Fan Speed temp threshold is set to 0
PR Number Synopsis Category: PTX10K Line Card specific interface PRs
1536270 Physical interface does not come back to up state after configuration changes at peer end.
Product-Group=evo 		If optics Rx LOS toggles from set to clear within a short interval of time, the interface may not able to lock to the new signal. This could happen in a number of scenarios. But primarily when there is config change on peer end, for example, IFD disable/enable, MTU change, LAG config, etc.
PR Number Synopsis Category: PTX10K specific platform PRs
1536497 The show chassis environment reports some ZF internal temperatures as 0 degrees C.
Product-Group=evo 		show chassis environment reports some ZF internal temperatures as 0 degrees C
PR Number Synopsis Category: Express PFE L2 fwding Features on ZX platforms
1540495 PTX10003: Multicast traffic drops after port is converted from trunk to access for a Layer 2 bridging case. Unicast and broadcast traffic are not affected.
Product-Group=evo 		When the IRB is configured in a bridge-domain, the multicast traffic drop might be seen after modifying the port configuration. The unicast and broadcast traffic is not affected.
 

20.3R1-S1-EVO - List of Known issues
PR Number Synopsis Category: AFT l2lam
1540129 [interface] [platformtag] : EVO-Brackla :: evo-aftmand core at L2almRouteHandler::notifyCommand after irb config commit with l2 config
Product-Group=evo 		With irb and L2 configuration and macs learned in vlans , evo-aftmand core can be seen when config is deleted.
PR Number Synopsis Category: Tracking PD specific issues on BRCM platforms running EVO
1531820 [pfe] [generic_evo] QFX5130-32CD :: JDI:RCT:IPCLOS:QFX5130-32CD:evo-pfemand Binding queue complete-deleted objects are seen with hwd::pcsE
Product-Group=evo 		QFX5130-32CD evo-pfemand Binding queue complete-deleted objects are seen with hwd::pcsE
PR Number Synopsis Category: EVO build infrastructure related issues.
1508626 Add Python 3.x for EVO - module list
Product-Group=evo 		There are Python 3.x modules that are missing from the Junos Evolved library. The equivalent of these modules is already installed under Python 2.7 module library.
1537947 New primary RPD might core in backup RE
Product-Group=evo 		In a scaled setup its possible that during GRES operation, the RPD on new primary RE can receive an ifstates update (DEL and CHANGE) and these are not expected during RPD re-syncing in such case the RPD can core and may recover automatically after sometime.
PR Number Synopsis Category: EVO Class of Services
1544531 [cos] [scheduler]: EVO COS: The COSd is not coming up online hence classifier/rewrite/scheduler functionalities are not working, after the Ingress FPC4 is restarted with 125% over-subs traffic (targeting all the interfaces in a BT chip (8x400G) on both ingress and egress sides)
Product-Group=evo 		 
PR Number Synopsis Category: Application controller related issues
1538626 With port mirror, the evo-pfemand may restarts unexpectedly on QFX platform running Junos Evolved.
Product-Group=evo 		With a port mirror function on a QFX platform running Junos Evolved, changing interface MTU may cause "evo-pfemand" process to restart unexpectedly.
1543421 [cos] [core]EVO COS: evo-aftmand-bt is cored at FPC5 (Lazurite LC) starting appc_get_next_object --> Controller::GetIncompleteGuidInfos --> NKDBNode::PopulateGuidInfo --> NKDB::ReasonStateCompressedCollisionPending --> evlruntime::evl_constructor, when the Ingress FPC4 (Scapa LC) is restarted with oversubs traffic (configs targeting all the interfaces in a BT chip (8x400G) on both ingress and egress sides)
Product-Group=evo 		 
PR Number Synopsis Category: software upgrade infra issues
1541509 PTX1008: Component Upgrade: Sw Rollback timed out when moving from image basedon 20.3R1 (mod in some apps) to20.3R1-202009111855.0-EVO same image
Product-Group=evo 		PTX1008: Component Upgrade: Sw Rollback timed out when moving from image basedon 20.3R1 (mod in some apps) to20.3R1-202009111855.0-EVO same image
1542064 PTX10008 : commit fails on re1 and succeeds on re0
Product-Group=evo 		PTX10008 : commit fails on re1 and succeeds on re0
PR Number Synopsis Category: EVO Netstack Juniper Tunnel Driver Module
1518609 [sysman] [generic_evo] Scapa : PDT - SCAPA - with graceful-restart enabled (isis/ldp/bgp), seeing packets loss during RE switch-over
Product-Group=evo 		EVO LDP graceful-restart in large topology may result in traffic loss.
PR Number Synopsis Category: EVO linux defects & enhancement requests
1528996 [interface] [interface] ptx10004 : PTX10004 :: The system login banner/ message is not showing up when we telnet to the device.
Product-Group=evo 		Standard "issue" message is not printed for the initial login attempt, but if that login attempt fails for any reason and the user is re-prompted, then the message is printed. This seems to be an issue common to PTX10004.
PR Number Synopsis Category: eventd, syslog infra issues
1545829 The syslog messages might not sent with correct port
Product-Group=evo 		If syslog port is configured to a port other than default one (UDP 514), the syslog messages are still sent to port 514. The syslog server might not receive the syslog messages so this issue has service impact.
PR Number Synopsis Category: SNMP, mib2d issues
1546784 Scapa : trace file is not created under snmp
Product-Group=evo 		Trace option is required to capture SNMP traps and also helps in debugging. It is not fixed in 20.3R1-S1-EVO.
PR Number Synopsis Category: Express PFE CoS Features
1543034 [cos] [scheduler] : Scapa : EVO COS: The "show interfaces voq ae2 non-zero" doesn't display the non-zero queue stats, where as the other AE bundle ae1 displays it correctly (both ae1 and ae2 has 4 x 400G member links each from the same BT chip)
Product-Group=evo 		To calculate AE VOQ stats, for every queue, FPC walks local AE legs and aggrerates VOQ stats. For each queue, the total number of non-zero PFEs for that queue is taken from the last leg. In the PR, AE1 had multiple legs. The ?last-walked? leg had no drops and therefore AE did not display VOQ drops.
1543067 [cos] [scheduler] : Scapa : EVO COS: The "show interfaces voq < ae/intf> non-zero" displays incorrect stats, once FPC comes up online after the offline/online/restart (seems "clear interfaces statistics all" doesnot clear this stats)
Product-Group=evo 		 
1543887 [cos] [scheduler] : Scapa : EVO COS: After clear interfaces statistics or after overnight run, the "show interfaces queue < ae1 / member links>" and "show interfaces < ae1/member links> extensive" commands display a bigger/magic number (negative wrap around value?), with 100% /95% line rate traffic is flowing thru the 8x400g and 4x100g Ingress --> to --> 8x400G Egress AE bundles ae1 and ae2, with the COS testing which targets stressing the full BT chip interfaces
Product-Group=evo 		 
1544551 [cos] [scheduler] : Scapa : EVO COS GRES: Consistent queue drops on high priority AF4 queue, after back-to-back GRES (by yank-out the RE0 followed by RE1) with 125% over-subs traffic (Customer configs targeting all the interfaces in a BT chip (8x400G) on both ingress and egress sides)
Product-Group=evo 		 
1545061 [cos] [scheduler] EVO COS: Queue drops on high priority AF4 queue is consistently seen, after one sib slot is offlined with 125% over-subs traffic (Customer configs targeting all the interfaces in a BT chip (8x400G) on both ingress and egress sides)
Product-Group=evo 		 
PR Number Synopsis Category: Express pfe ddos protection feature
1529847 Fpc reboots if ddos violations are seen during fpc restart
Product-Group=evo 		If ddos violations are seen during fpc restart, crash is seen and fpc restarts
PR Number Synopsis Category: Express PFE FW Features
1540625 PTX10008 : Core generated while testing AE filters
Product-Group=evo 		PTX10008 : Core generated while testing AE filters
PR Number Synopsis Category: Express PFE L3 Multicast
1513474 Out of order delete messages are seen when doing AE link flap
Product-Group=evo 		On a large scale routing on PTX10008/PTX10016, when an AE interface flaps, during next-hop convergence, you may see syslog filled with "Possible out of order deleting of AftNode" error messages.
1531859 Scapa : Multicast traffic loss seen on reciever port when one of the FPC with AE member link is brought offline
Product-Group=evo 		When multicast traffic is running over AE bundle, which spans multiple FPCs, if one of FPCs is brought offline, the traffic on that FPC will be impacted. it's expected to see some traffic drops.
PR Number Synopsis Category: PRs for Lazurite Platform and Interface
1538514 [mgdtag] [mgdtag] : [lazurite]: After config "global system name-server" config commit should fail but commit is succeeds.
Product-Group=evo 		 
PR Number Synopsis Category: RPD Interfaces related issues
1542663 [cos] [core] :: EVO COS [20.3R1.14-EVO]: rpd cores after the ingress FPC4 is restarted while 100% line rate traffic is flowing thru the 8x400G ingress --> to --> 8x400G Egress AE bundles, with COS testing which targets stressing the full BT chip interfaces
Product-Group=evo 		 
PR Number Synopsis Category: KRT Queue issues within RPD
1511539 PTX10008 EVO packet loss around 1% was seen from Spirent and Regular discard increase (show pfe statistics traffic) after GRES check pass and then GRES
Product-Group=evo 		PTX10008 EVO packet loss around 1% was seen from Spirent and Regular discard increase (show pfe statistics traffic) after GRES check pass and then GRES
PR Number Synopsis Category: show route table commands, tracing, and syslog facilities
1537546 [eoam] [eoamtag] ACX-R6675 :: JDI-ACCESS_REGRESSION: ACX-R6675 : rpd core seen at l2ckt_vc_adv_recv, l2ckt_adv_rt_flash (taskptr=0x4363b80, rtt=0x4418100, rtl=< optimized out>, data=< optimized out>, opcode=< optimized out>) at ../../../../../../../../../src/junos/usr.sbin/rpd/l2vpn/l2ckt.c:7982
Product-Group=evo 		Trigger: ---------- -Issue is seen only on arm64 based platform. -Issue is seen we have enabled l2circuit trace log. -Have basic l2circuit active neighbour configuration . Symtom: ----------- -It generates a RPD core. Workaround: --------------- Do not enable l2circuit trace log.
PR Number Synopsis Category: PTX10K Line Card specific interface PRs
1542681 PtX10008 JNP10K-LC1201 : One of the ingress side 400G interfaces keeps flapping, when specific COS config is loaded with the 20.3R1.14-EVO image (The config uses all the 8x400G interfaces from the same BT chip. i.e 8x400G ingress traffic going to 8x400G Egress AE bundles)
Product-Group=evo 		PtX10008 JNP10K-LC1201 : One of the ingress side 400G interfaces keeps flapping, when specific COS config is loaded with the 20.3R1.14-EVO image (The config uses all the 8x400G interfaces from the same BT chip. i.e 8x400G ingress traffic going to 8x400G Egress AE bundles)
PR Number Synopsis Category: PTX10K specific platform PRs
1543428 [cos] [core] : Scapa : EVO COS: The hwdre is cored starting __cxxabiv1::__cxa_throw --> __cxxabiv1::__terminate --> __gnu_cxx::__verbose_terminate_handler --> __GI_raise, resulting in queue drops on high priority queue AF4, after back-to-back GRES with oversubs traffic w/ COS configs targeting all the interfaces in a BT chip (8x400G) on both ingress and egress sides
Product-Group=evo 		 
PR Number Synopsis Category: PTX10K RE EVO Issues
1496895 The copying of files to the RCB over WAN ports is slow. This is observed across all platforms running Junos OS Evolved platforms.
Product-Group=evo 		The copying of files to the RCB over WAN ports is slow. This is observed across all platforms running Junos OS Evolved platforms.
 

Modification History

First publication 2020-11-11