CSCwr72707

On fully populated Cisco UCS C220 M8, C240 M8, or UCSXE130C M8 nodes with more than two Kioxia E3.S CD8P 1.9 T NVMe drives, random workload performance drops significantly during concurrent tests. Specifically, 4K random read performance may drop from a rated 1,600kIOPS per drive to approximately 680k Input/Output Operations Per Second (IOPS) per drive during concurrent operations. Single drive tests meet expected performance, indicating the issue only occurs with multi-drive setups

To resolve this performance bottleneck, apply the following Intel-recommended kernel tuning parameters in the Red Hat Enterprise Linux (RHEL) boot configuration:

• For single-drive configurations: Add the following to the RHEL kernel command line:

  intel_pstate=disable

  pcie_aspm.policy=performance

  pci=pcie_bus_safe

• For multi-drive (fully populated) configurations: Add the following to the RHEL kernel command line:

  intel_pstate=disable

  pcie_aspm.policy=performance

  pci=pcie_bus_safe

  nvme.poll_queues=32

  intel_iommu=on

X-series: 6.0(2.260040)

XE-series: 6.0(2.260042)

C-series: 6.0(2.260044)

B-series: 6.0(2.260040)

CSCws25068

When installing Ubuntu 24.04.3 with kernel version 6.8.0-71 on the Unified Edge server, the integrated network card (ICE driver) is not detected, preventing network configuration and connectivity. This issue occurs because kernel 6.8.0-71 lacks ICE driver support for the NIC on this platform.

1. Download the GA ISO from the Ubuntu Noble Release Page and install it using a bootable USB or virtual media

2. Download the required 6.8.0-80 kernel packages (.deb files) from the Ubuntu Security Pool.

3. Place all .deb files (linux-image, linux-headers, linux-modules, linux-modules-extra, linux-tools) in a folder and create an ISO using a tool like genisoimage or mkisofs.

4. Mount the ISO on the server after installing Ubuntu.

5. Install the kernel packages using dpkg -i *.deb.

6. Reboot the server and verify the kernel version is 6.8.0-80.

The integrated network card will now be detected and available for configuration.

6.0(1.251030)

CSCwn95239

On Cisco UCS C-Series servers, the server may fail to power on due to BIOS corruption. The following message is displayed: "Power on failure asserted [Power on Fail, Server failed to power-on]."

This issue can occur in the following scenarios, especially when BIOS update operations are disrupted:

• Interrupting a BIOS flash or update (for example, forcing a power cycle during the update).

• Initiating a CMOS clear operation during a BIOS update.

• Restoring factory defaults during a BIOS update.

• Rebooting Cisco IMC during a BIOS update.

• Performing other unexpected operations that disrupt the BIOS update process.

Do one of these:

• Reboot of CIMC. In-turn, recovery mechanism will start that is bios activation.

• Activate of bacup BIOS.

• Update and activate BIOS.

4.3(6.250039)

CSCwo68248

On Cisco UCS C220 M7 and C240 M7 servers running HUU 4.3(3.240xxx) level firmware with 2 PCIe VICs and 1 MLOM VIC, upgrading the server firmware to HUU 4.3(4.xxxxxx), 4.3(5.xxxxxx), 4.3(6.xxxxx), or later using the Redfish API interface may cause the UCSUpdate task to falsely report an update failure for the PCIe VIC (riser1) through which CiscoCard mode is configured. However, the firmware is actually updated and activated successfully in the background.

There is no workaround. Ignore the false failure update.

4.3(3.240043)