Release Notes
Table Of Contents
- Dell EMC PowerEdge Systems Running Red Hat Enterprise Linux 8 Release Notes
- Contents
- Release summary
- Compatibility
- New and enhanced in RHEL 8 release
- Important notes
- Fixes
- The mcelog utility logs 'only decoding architectural errors' message in var/log/messages
- Disk drives part of MD RAID are not listed as installation destination by the installer
- Dell EMC OpenManage Storage Services utility fails to reconfigure the virtual disk
- Guest VMs with SRIOV VFs assigned take a long time to power on, and libvirt related errors are observed
- Dmesg displays Integrity Measurement Architecture (IMA) driver related-messages during system boot
- After every reboot, the network interface name changes
- Red Hat Enterprise Linux Version 8 installation wizard creates a duplicate bonding interface
- Servers with the AMD Rome processor display a CCP initialization failure message in dmesg
- PowerEdge servers with the AMD Rome processor fail to detect an NVMe drive after multiple hot plugs
- Operating system enters the dracut shell during boot
- System crashes when rebooted with SR-IOV-enabled QLogic cards
- After system reboot, Disk data format (DDF) devices are not listed in /proc/mdstat
- Updating NVMe firmware using the nvme-cli utility displays an error in dmesg
- Fatal error BDF 02:00.0 is detected with BCM574xx NICs
- NVMe devices are not detected after hot-plugging
- Linux operating system fails to detect the Intel x710 card
- Dmidecode displays OUT OF SPEC in Slot Type and Slot Length of SMBIOS system slots
- Custom partitioning fails with FC LUN
- When booting the system from iSCSI with Mellanox CX-4 and CX-5 adapters, the system reports csum failure message
- Red Hat Enterprise Linux 8 kernel panic is observed due to fatal hardware error
- Known issues
- Operating system crashes on systems with NVIDIA GPGPUs
- Dmesg and /var/log/messages display AMD-Vi related messages
- The status of the NetworkManager service may be inactive when RHEL 8.3 is rebooted
- Operating system crashes on AMD Rome CPU-based systems and with Intel E810 NIC
- The lvcreate command requests a response from the user when -wipesignature -yes parameters are passed
- The mdmonitor service displays an error during operating system installation
- The dmidecode utility displays the slot type as
for PCIe Gen 4 NVMe slots - NetworkManager may restart unexpectedly when creating greater than 256 VLAN devices configured with DHCP IP
- FCoE session is not reestablished after MX9116N switch is rebooted
- Dmesg shows drm related call trace in RHEL 8.3
- Dmesg displays error messages when NVMe device is surprise removed
- Status of the RAID 0 logical volume is displayed as Available when one of the members of the RAID array is surprise removed
- /proc/mdstat and mdadm -D commands display incorrect statuses when two NVMe devices are surprise removed from a RAID 5 MD array
- Dell Controlled Turbo feature is not functional
- Caps Lock key-press is not registered on the Dell PowerEdge iDRAC virtual console
- RHEL 8.3 installer does not automatically locate the source installation repository when only inst.stage2=hd boot option is used
- The output of the systemctl status command displays the status as thawing
- Advanced Configuration and Power Interface (ACPI) error messages displayed in dmesg
- Drivers available in OEMDRV drive are not installed during the operating system installation
- The Mellanox IB devices are listed under an incorrect device category on Red Hat Enterprise Linux 8
- The lspci utility is unable to read Vital Product Data (VPD) from QLogic QLE2692 adapter
- Driver dependency mismatch errors reported while installing out-of-box drivers on Red Hat Enterprise Linux 8.x
- Dmesg displays TPM and nvdimm related-messages in Red Hat Enterprise Linux 8.1
- Link Up message is observed when the NVMe device slot is powered off and the device is unplugged
- Mellanox InfiniBand adapters are listed in Bluetooth
- iscsiadm output displays STATIC in the iface.bootproto field when the network interface is configured to DHCP
- When system reboots, system stops responding at the end of the reboot process
- Unable to shut down RHEL 8 when you select Graceful shutdown option or when you press power button on the server
- RHEL 8 does not discover FCoE LUNs connected over Broadcom BCM57XXX NICs
- iSCSI LUN not discovered during RHEL 8 installation
- RHEL 8 installation fails on systems with Emulex OneConnect card
- Switching between runlevels fails
- Limitations
- Resources and support
- Contacting Dell EMC
NetworkManager may restart unexpectedly when
creating greater than 256 VLAN devices configured
with DHCP IP
Description: On RHEL 8.3, when more than 256 VLAN devices are created and configured with DHCP IP,
NetworkManager utility may restart unexpectedly resulting in not creating expected number of VLAN
devices and some VLAN interfaces may not receive DHCP IP.
Applies to: Red Hat Enterprise Linux 8.0 and later
Workaround: 1. Increase the limit on number of files that NetworkManager can open as below:
a. Create file/etc/systemd/system/NetworkManager.service.d/set-limit-nofile.conf
b. Add following lines to /etc/systemd/system/NetworkManager.service.d/set-limit-nofile.conf
[Service] LimitNOFILE=65536
c. systemctl daemon-reload
d. systemctl restart NetworkManager or reboot the system after step (b).
2. Disable IPv6 configuration for the VLAN interfaces at the time of connection activation. Instead of
disabling IPv6 in two steps, disable it in a single step as shown below.
a. Merge steps b) and c) to a single step d).
b. nmcli con add type vlan ifname <interface name> con-name <connection name> dev <device
name> id <vlan id>
c. nmcli con modify <connection name> ipv6.method disable
d. nmcli con add type vlan ifname <interface name> con-name <connection name> dev <device
name> id <vlan id> ipv6.method disabled
Systems
affected:
All Dell EMC PowerEdge systems
Tracking number: 191829
FCoE session is not reestablished after MX9116N
switch is rebooted
Description:
On Dell EMC PowerEdge MX740C system containing Qlogic QL41000 series CNA device configured for
FCoE, FCoE session is not restored automatically when the MX9116N switch is rebooted.
Applies to: Red Hat Enterprise Linux 8.2
Cause: Issue in qedf driver related to handling the switch reboot scenario.
Workaround: To reestablish FCoE session execute the command:
rescan-scsi-bus.sh -i --hosts=<host number>
The <host number> refers to the host number associated with the lost FCoE session and this can be
identified from dmesg.
Systems
affected:
All Dell EMC systems supporting Qlogic QL41000 series CNA devices.
Tracking number: 190870
Known issues 21