Merge pull request #3056 from CumulusNetworks/create-pull-request/patch

Changes by create-pull-request action

Author: mergify[bot]

content/cumulus-linux-511/Whats-New/rn.md

@@ -355,8 +355,8 @@ pdfhidden: True
 | <a name="4521203"></a> [4521203](#4521203) <a name="4521203"></a> <br /> | The SATA driver on the NVIDIA SN2410 switch sometimes exhibits failures similar to the following:<br><pre>2025-06-20T20:03:38.785966+09:00 S1-RF10B kernel: <br>&#91;49878794.456350&#93; ata1.00: failed command: WRITE FPDMA QUEUED<br>2025-06-20T20:03:38.785968+09:00 S1-RF10B kernel: &#91;49878794.461984&#93; ata1.00: cmd<br>61/08:08:50:2b:c4/00:00:01:00:00/40 tag 1 ncq dma 4096 out<br>2025-06-20T20:03:38.785970+09:00 S1-RF10B kernel: &#91;49878794.461984&#93; res 40/00:0c:50:2b:c4/00:00:01:00:00/40 Emask 0×10 (ATA bus error)</pre><br>To work around this issue, power cycle the switch. | 5.5.1-5.13.1 | 5.14.0-5.15.1|
 | <a name="4508830"></a> [4508830](#4508830) <a name="4508830"></a> <br /> | Cumulus Linux allows you to add bond ports of mismatched speeds (such as 10G and 25G) to the same LACP bond without error and the bond reports UP.  | 5.11.2-5.15.1 | |
 | <a name="4507163"></a> [4507163](#4507163) <a name="4507163"></a> <br /> | In some cases after a package upgrade with ISSU (warm boot mode), you see continuous errors in the syslog similar to the following:<br><pre>NOTICE  CORE_ASYNC: Error at pre send callback status &#91;Resource is in use&#93;</pre> | 5.11.2-5.11.3 | 5.11.4-5.15.1|
-| <a name="4499025"></a> [4499025](#4499025) <a name="4499025"></a> <br /> | You see a high volume of NAT and NFCT errors flooding <code>switchd</code> logs. This issue has no functional impact. | 5.11.2-5.15.1 | |
 | <a name="4495231"></a> [4495231](#4495231) <a name="4495231"></a> <br /> | If the hardware clock date is later than 2038-01-19 03:14:07 UTC, the image might fail to install due to a grub-install failure on the EFI filesystem, which is a VFAT filesystem.  As a result, you see the grub prompt immediately after reboot.  To work around this issue, reboot from the grub prompt to go into ONIE.  From ONIE, use the date command to set a date before 2038-01-19, then run the <code>hwclock --systohc</code> command to add it to the hardware clock.  You can then use <code>onie-nos-install</code> to install the image. | 5.9.0-5.15.1 | |
+| <a name="4486668"></a> [4486668](#4486668) <a name="4486668"></a> <br /> | You see a high volume of NAT and NFCT errors flooding <code>switchd</code> logs. This issue has no functional impact. | 5.11.2-5.13.1 | 5.14.0-5.15.1|
 | <a name="4486200"></a> [4486200](#4486200) <a name="4486200"></a> <br /> | If you enable dynamic NAT and try to install two identical dynamic NAT rules, <code>switchd</code> might crash. | 5.11.2-5.13.1 | 5.14.0-5.15.1|
 | <a name="4475111"></a> [4475111](#4475111) <a name="4475111"></a> <br /> | When you try to convert a layer 3 port that is part of ECMP to a bond member,  you might see a failure in the <code>switchd</code> logs. This issue does not have any functional impact. | 5.11.2-5.15.1 | 5.9.4|
 | <a name="4472414"></a> [4472414](#4472414) <a name="4472414"></a> <br /> | After you modify the IP address of an SVI, DHCP relay uses an old cached IP address instead of the changed IP address. This occurs because DHCP relay monitors the link state change not IP address change. | 5.11.0-5.14.0 | 5.15.0-5.15.1|
@@ -481,9 +481,9 @@ pdfhidden: True
 | <a name="4501645"></a> [4501645](#4501645) <a name="4501645"></a> <br /> | When you connect the QLOUIE cable with FW 52.181.1003 to an NVIDIA SN5600 switch, you might see link flaps.  | 5.11.0-5.11.1 | |
 | <a name="4501637"></a> [4501637](#4501637) <a name="4501637"></a> <br /> | If you configure TACACS+ authentication and you restrict certain TACACS+ users from running specific commands, after you upgrade the <code>libtacplus-map1</code> package with the <code>dpkg</code> command or with package upgrade, authorization for those users might no longer work correctly (due to permission issues in their home directories). To resolve this issue, run the <code>ls -l /home/</code> command to determine which <code>tacacs*</code> home directories the <code>bin</code> user owns and make sure the directories have 0755 (drwxr-xr-x) permissions. |  | |
 | <a name="4495502"></a> [4495502](#4495502) <a name="4495502"></a> <br /> | After you run the <code>onie-install -t</code> command to upgrade the switch, accounting does not work but authentication does. This issue occurs because <code>rsyslog.service</code> does not restart properly after upgrade. To work around this issue, reboot the switch a second time or manually restart <code>rsyslog.service</code> with the <code>systemctl restart rsyslog.service</code> command. |  | |
-| <a name="4475117"></a> [4475117](#4475117) <a name="4475117"></a> <br /> | After you perform an optimized, two partition upgrade followed by a rollback to the previous release, certain services, including the NVUE service, might fail to start. If you see this issue after rollback, run the following commands on the switch to ensure that file ownership is correct:<br><pre>sudo chown -R nvue /var/lib/nvue<br>sudo chown -R ntpsec /var/lib/ntpsec<br>sudo chown -R Debian-snmp /var/lib/snmp</pre><br>To restart the services, either reboot the switch, or run the following commands:<br><pre>sudo systemctl restart nvued.service<br>sudo systemctl restart nvue-startup.servicesystemctl is-enabled snmpd && sudo systemctl restart snmpd.servicesystemctl is-enabled ntpsec && sudo systemctl restart ntpsec.service</pre> | 5.11.1 | |
 | <a name="4475074"></a> [4475074](#4475074) <a name="4475074"></a> <br /> | The SN5610 switch records a High FEC Bin Error at room temperature. | 5.13.0-5.13.1 | |
 | <a name="4472549"></a> [4472549](#4472549) <a name="4472549"></a> <br /> | When running tens of thousands of <code>nv set</code> commands, the <code>/var/lib/nvue</code> directory might grow to several GBs in size, potentially using all the disk space. To work around this issue, run the following commands to reduce the disk space in the <code>/var/lib/nvue</code> directory:<br><pre>cumulus&#64;switch:~$ sudo su<br>cumulus&#64;switch:~$ cd /var/lib/nvue/config<br>cumulus&#64;switch:~$ git gc</pre> | 5.13.0-5.13.1 | |
+| <a name="4472217"></a> [4472217](#4472217) <a name="4472217"></a> <br /> | After you perform an optimized, two partition upgrade followed by a rollback to the previous release, certain services, including the NVUE service, might fail to start. If you see this issue after rollback, run the following commands on the switch to ensure that file ownership is correct:<br><pre>sudo chown -R nvue /var/lib/nvue<br>sudo chown -R ntpsec /var/lib/ntpsec<br>sudo chown -R Debian-snmp /var/lib/snmp</pre><br>To restart the services, either reboot the switch, or run the following commands:<br><pre>sudo systemctl restart nvued.service<br>sudo systemctl restart nvue-startup.servicesystemctl is-enabled snmpd && sudo systemctl restart snmpd.servicesystemctl is-enabled ntpsec && sudo systemctl restart ntpsec.service</pre> | 5.11.1 | |
 | <a name="4458897"></a> [4458897](#4458897) <a name="4458897"></a> <br /> | The <code>nv config patch</code> and <code>nv config replace</code> commands have no effect with filenames with relative paths, such as <code>./config.yaml</code>, <code>../config.yaml},  and so on. | 5.13.0-5.13.1 | |
 | <a name="4457389"></a> [4457389](#4457389) <a name="4457389"></a> <br /> | On rare occasions, when bridge or L2VNI interfaces are coming up or transitioning state, type 2 EVPN routes might not be properly installed. To work around this issue, flap the VNI interface. | 5.9.3 | |
 | <a name="4423430"></a> [4423430](#4423430) <a name="4423430"></a> <br /> | When toggling the bridge binding flag on an SVI from ON to OFF, the SVI might not come operationally UP if it was DOWN previously from the bridge binding flag. | 5.11.0-5.11.1 | |
@@ -511,7 +511,7 @@ pdfhidden: True
 | <a name="4423134"></a> [4423134](#4423134) <a name="4423134"></a> <br /> | On Spectrum-1a switches with IGMP snooping enabled, a multicast hardware programming failure might occur after interface flap or switch reboot events. You can observe this issue when log messages similar to the following are generated:<br><pre>sx_sdk&#91;18174&#93;: ERROR   FDB: Usage API type can not be changed for 0x1003 fid.<br>sx_sdk&#91;18174&#93;: ERROR   FDB: Failed to __fdb_unreg_mc_flood_cfg_api_type_set , err: Command Unpermitted<br>switchd&#91;19460&#93;: hal_mlx_l2mc.c:1107 ERR VFID: 4099, Failed to set unregistered IPv4 MC mode FLOOD and attr MCC 0: Command Unpermitted</pre> | 5.12.0-5.12.1 | |
 | <a name="4422898"></a> [4422898](#4422898) <a name="4422898"></a> <br /> | When displaying BGP neighbor advertised routes in JSON format with very large routing tables, the process consumes excessive memory resulting in out of memory crashes.  | 5.0.0-5.11.1, 5.13.0-5.13.1 | |
 | <a name="4408280"></a> [4408280](#4408280) <a name="4408280"></a> <br /> | NVUE commands create excessive log data. To work around this issue, configure rsyslog rules to limit logging of these commands. | 5.11.0-5.11.1, 5.12.0-5.12.1 | |
-| <a name="4408161"></a> [4408161](#4408161) <a name="4408161"></a> <br /> | On the NVIDIA SN2201 switch, the fan tray LED status update fails and you see the following syslog errors:<br><pre>systemd-udevd116276: mlxreg:fan1:green: Process ‘/usr/bin/hw-management-chassis-events.sh fantray-led-event mlxreg:fan1:green 255’ failed with exit code 1.</pre><br>To work around this issue, restart the <code>hw-management</code> service with the <code>sudo systemctl restart hw-management</code> command. | 5.7.0-5.9.3 | |
+| <a name="4404955"></a> [4404955](#4404955) <a name="4404955"></a> <br /> | On the NVIDIA SN2201 switch, the fan tray LED status update fails and you see the following syslog errors:<br><pre>systemd-udevd116276: mlxreg:fan1:green: Process ‘/usr/bin/hw-management-chassis-events.sh fantray-led-event mlxreg:fan1:green 255’ failed with exit code 1.</pre><br>To work around this issue, restart the <code>hw-management</code> service with the <code>sudo systemctl restart hw-management</code> command. | 5.7.0-5.9.3 | |
 | <a name="4400384"></a> [4400384](#4400384) <a name="4400384"></a> <br /> | After the switch reboots or <code>switchd.service</code> restarts, NVUE applied ERSPAN sessions do not work if the ERSPAN destination IP address is reachable through an MLAG bond. To work around this issue, remove the ERSPAN configuration and reapply it using NVUE. | 5.11.0-5.11.1 | |
 | <a name="4389124"></a> [4389124](#4389124) <a name="4389124"></a> <br /> | NVUE fails when applying ERSPAN configuration on an MLAG peer. This failure occurs because ERSPAN does not support bond slave ports as analyzer ports but fails to validate the configuration.  | 5.11.0-5.11.1 | |
 | <a name="4370955"></a> [4370955](#4370955) <a name="4370955"></a> <br /> | After enabling, then disabling truncation on a SPAN session,  truncated packets are still received on the SPAN destination. To work around this issue, remove the SPAN session configuration, reboot the switch, then reconfigure the SPAN session without truncation. | 5.12.0-5.12.1 | |
@@ -549,8 +549,8 @@ pdfhidden: True
 | <a name="4521203"></a> [4521203](#4521203) <a name="4521203"></a> <br /> | The SATA driver on the NVIDIA SN2410 switch sometimes exhibits failures similar to the following:<br><pre>2025-06-20T20:03:38.785966+09:00 S1-RF10B kernel: <br>&#91;49878794.456350&#93; ata1.00: failed command: WRITE FPDMA QUEUED<br>2025-06-20T20:03:38.785968+09:00 S1-RF10B kernel: &#91;49878794.461984&#93; ata1.00: cmd<br>61/08:08:50:2b:c4/00:00:01:00:00/40 tag 1 ncq dma 4096 out<br>2025-06-20T20:03:38.785970+09:00 S1-RF10B kernel: &#91;49878794.461984&#93; res 40/00:0c:50:2b:c4/00:00:01:00:00/40 Emask 0×10 (ATA bus error)</pre><br>To work around this issue, power cycle the switch. | 5.5.1-5.13.1 | 5.14.0-5.15.1|
 | <a name="4501645"></a> [4501645](#4501645) <a name="4501645"></a> <br /> | When you connect the QLOUIE cable with FW 52.181.1003 to an NVIDIA SN5600 switch, you might see link flaps.  | 5.11.0-5.13.1 | 5.14.0-5.15.1|
 | <a name="4495231"></a> [4495231](#4495231) <a name="4495231"></a> <br /> | If the hardware clock date is later than 2038-01-19 03:14:07 UTC, the image might fail to install due to a grub-install failure on the EFI filesystem, which is a VFAT filesystem.  As a result, you see the grub prompt immediately after reboot.  To work around this issue, reboot from the grub prompt to go into ONIE.  From ONIE, use the date command to set a date before 2038-01-19, then run the <code>hwclock --systohc</code> command to add it to the hardware clock.  You can then use <code>onie-nos-install</code> to install the image. | 5.9.0-5.15.1 | |
-| <a name="4475117"></a> [4475117](#4475117) <a name="4475117"></a> <br /> | After you perform an optimized, two partition upgrade followed by a rollback to the previous release, certain services, including the NVUE service, might fail to start. If you see this issue after rollback, run the following commands on the switch to ensure that file ownership is correct:<br><pre>sudo chown -R nvue /var/lib/nvue<br>sudo chown -R ntpsec /var/lib/ntpsec<br>sudo chown -R Debian-snmp /var/lib/snmp</pre><br>To restart the services, either reboot the switch, or run the following commands:<br><pre>sudo systemctl restart nvued.service<br>sudo systemctl restart nvue-startup.servicesystemctl is-enabled snmpd && sudo systemctl restart snmpd.servicesystemctl is-enabled ntpsec && sudo systemctl restart ntpsec.service</pre> | 5.11.1 | 5.11.2-5.15.1|
 | <a name="4472414"></a> [4472414](#4472414) <a name="4472414"></a> <br /> | After you modify the IP address of an SVI, DHCP relay uses an old cached IP address instead of the changed IP address. This occurs because DHCP relay monitors the link state change not IP address change. | 5.11.0-5.14.0 | 5.15.0-5.15.1|
+| <a name="4472217"></a> [4472217](#4472217) <a name="4472217"></a> <br /> | After you perform an optimized, two partition upgrade followed by a rollback to the previous release, certain services, including the NVUE service, might fail to start. If you see this issue after rollback, run the following commands on the switch to ensure that file ownership is correct:<br><pre>sudo chown -R nvue /var/lib/nvue<br>sudo chown -R ntpsec /var/lib/ntpsec<br>sudo chown -R Debian-snmp /var/lib/snmp</pre><br>To restart the services, either reboot the switch, or run the following commands:<br><pre>sudo systemctl restart nvued.service<br>sudo systemctl restart nvue-startup.servicesystemctl is-enabled snmpd && sudo systemctl restart snmpd.servicesystemctl is-enabled ntpsec && sudo systemctl restart ntpsec.service</pre> | 5.11.1 | 5.11.2-5.15.1, 5.14.0-5.15.1|
 | <a name="4461102"></a> [4461102](#4461102) <a name="4461102"></a> <br /> | In certain cases, when a port is down and you apply adaptive routing with the link utilization threshold setting to the port as it goes up, you might see log errors while the port is not yet up. | 5.11.0-5.13.1 | 5.14.0-5.15.1|
 | <a name="4423430"></a> [4423430](#4423430) <a name="4423430"></a> <br /> | When toggling the bridge binding flag on an SVI from ON to OFF, the SVI might not come operationally UP if it was DOWN previously from the bridge binding flag. | 5.11.0-5.12.1 | 5.13.0-5.15.1|
 | <a name="4423365"></a> [4423365](#4423365) <a name="4423365"></a> <br /> | When you enable NAT dynamic mode and NAT rules with NVUE in a single commit, you see the error <code>error: hw sync failed (Dynamic NAT is not enabled. Ignoring rules..)</code>. To work around this issue, enable dynamic NAT with NVUE in one commit, then add NAT ACL rules in a subsequent commit. | 5.11.0-5.11.4 | 5.12.0-5.15.1|