Known Issues

Installation and Upgrade Issues

The following are known installation and upgrade issues that have been encountered in this release of Oracle Linux 8.

Installer automatically enables Ethernet over USB network interface during a PXE installation

During a Preboot Execution Environment (PXE) installation of Oracle Linux 8, the installer automatically enables the Ethernet over USB network interface with the bootproto=dhcp and ONBOOT=yes parameters. These default settings causes the NetworkManager service to fail to start.

To prevent this issue from occurring, or to resolve the issue if you have already encountered it, use one of the following workarounds:

Prior to installation, disable the ONBOOT parameter for the Ethernet over USB network interface in the kickstart file, as follows:
```
network --bootproto=dhcp --device=enp0s20f0u8u3c2 --onboot=off --ipv6=auto
```
During installation, on the Network & Host Name screen, do not select the Connect automatically with priority check box to connect automatically on every reboot for the Ethernet over USB network interface.
If you have already encountered this issue, then after the installation, change the network configuration setting for the Ethernet over USB network interface to ONBOOT=no. Then reboot the system.

(Bug ID 31888490)

Interactive text-based installation wizard unable to complete when an alternate language is selected

If you selected an alternate language while using the text-based installer to install the OS, you cannot proceed with the installation. The installation is blocked with [!] flags for Software Selection and Installation Destination regardless of what you have set for these two options.

However, this issue does not occur if you are performing an installation by using the default English language selection or by using the graphical installation program.

(Bug IDs 30535416, 29648703)

Changing installation source results in errors if alternative installation repository is set at boot

If the installer is booted with the inst.repo option set, changing the installation source to use a CD or DVD device within the installer results in an error that prevents you from continuing the installation, unless you set the source back to the original source that was set at boot.

If you set the inst.repo option to point to a hard disk and then attempt to change the installation source inside the installer, the installer displays an error; but, you can still proceed with the installation.

To avoid these issues, do not set the inst.repo option at boot if you do not intend to use the installation source that is provided. Or, use the inst.repo source that is defined at boot without attempting to change installation source inside the installer.

(Bug ID 30316179)

Graphical installation program fails to produce error when an unacceptable Kdump value is entered

A minor upstream usability error affects the graphical installation program during the configuration of Kdump.

If you specify an unacceptable value when manually configuring the Kdump memory reservation, you can click Done to return to the Installation Summary screen. The installer does not generate a warning or error message. Instead, the installer automatically resets the value either to the last known acceptable value or the default value of 512 MB, which enables the installation to succeed. However, because this corrected setting is not displayed on the screen, you might not become aware that your specified value was ignored.

This issue does not occur with the text-based installer, which correctly returns an error if you enter an unacceptable value and prevents you from continuing.

(Bug IDs 31133351, 31182708)

Graphical installation program does not display the reserved memory that is manually set for Kdump

A minor usability error affects the graphical installation program during the configuration of Kdump. If you manually change the default memory size that is reserved for Kdump, the new setting is not displayed when the screen is refreshed. Instead, only the values for the total system memory and usable system memory are displayed. Consequently, the limits for the parameter "Memory to be reserved (Mb)" become unknown for future Kdump configuration.

Note:

The default setting auto for Kdump memory reservation is adequate as the kernel determines what size to use when it boots

(Bug IDs 31133287 and 31182699)

Scriptlet-related error for microcode_ctl might be displayed during upgrade

A scriplet-related error message might be displayed during an upgrade of an Oracle Linux 8 release to its next version. When you run the dnf update command, an output similar to the following might appear:

  Running scriptlet: tuned-2.13.0-6.0.2.el8.noarch                            
             1089/1089
  Running scriptlet: microcode_ctl-4:20191115-4.el8.x86_64                    
             1089/1089
realpath: weak-updates/kmod-kvdo/vdo/kvdo.ko: No such file or directory
realpath: weak-updates/kmod-kvdo/uds/uds.ko: No such file or directory
dracut: installkernel failed in module kernel-modules-extra
warning: %posttrans(microcode_ctl-4:20191115-4.el8.x86_64) scriptlet failed,
exit status 1

Error in POSTTRANS scriptlet in rpm package microcode_ctl
  Running scriptlet: libgcc-8.3.1-4.5.0.7.el8.x86_64                          
             1089/1089
  Running scriptlet: glibc-common-2.28-101.0.1.el8.x86_64                    
             1089/1089
  Running scriptlet: info-6.5-6.el8.x86_64                                    
             1089/1089

This error message is displayed if you use the Server with GUI environment to install Oracle Linux 8 and then you reboot the server by using RHCK. This installation method installs the kernel dependent, kmod-kvdo package or module, which is a different version in the previous Oracle Linux 8 release.

However, you can safely ignore the message because the kmod-kvdo package is successfully installed during the upgrade process.

Note:

This error does not occur if you install the Minimal Install base environment or if you boot the server with UEK R6 or UEK R7.

(Bug ID 31292199)

Version of libpcap packages in BaseOS channel does not support RDMA

The libpcap packages that are in the BaseOS channel do not support dumping RDMA sniffer information. After installing or upgrading to Oracle Linux 8.3, if you require RDMA functionality, you should run the dnf downgrade command to downgrade your libpcap version so that you can use the libpcap packages that are published to the ol8_UEKR6_RDMA repository, for example:

sudo dnf downgrade libpcap

(Bug ID 32049290)

rhnreg_ks register command might fail if python3-rhn-virtualization-host package is installed

Beginning with Oracle Linux 8.1, using the rhnreg_ks command to register a system with the Unbreakable Linux Network (ULN)might fail if the python3-rhn-virtualization-hosts package is installed on the system. This issue has been observed when the libvirtd service is not running.

To work around this issue, ensure that the libvirtd packages are installed on your system and that the service is enabled and running prior to issuing the rhnreg_ks command.

(Bug ID 30366521)

Presence of beignet package could result in dependency issue during an upgrade

While upgrading a system to the current Oracle Linux 8 release, you might encounter a dependency issue if the beignet package exists on the system to be upgraded.

This issue exists specifically in cases where you upgrade systems running Oracle Linux 8.2 or earlier releases to the current Oracle Linux version. In these earlier releases, the beginet package requires earlier versions of the clang-libs package.

However, the beignet package is currently not available for Oracle Linux 8.4 and later Oracle Linux 8 releases. Therefore, the issue does not exist for these cases.

To work around this issue, remove the beignet package from the system prior to upgrading to the current Oracle Linux 8 release.

(Bug ID 31213935)

ULN registration wizard not displayed on first boot after an installation

On new installations of Oracle Linux 8, the ULN registration wizard that presents the options to register with ULN and to use Oracle Ksplice is not displayed on first boot.

As an alternative, you can register with ULN after the installation completes. For instructions, see https://linux.oracle.com/.

(Bug ID 29933974)

Graphics controller requirements for an installation on an Oracle VM VirtualBox guest

To successfully install Oracle Linux 8 on an Oracle VM VirtualBox guest, where the graphical installation program is used and the default Server with GUI environment is selected, you must set the guest to use the VMSVGA graphics controller and configure the guest with at least 64MB of memory. Otherwise, the graphical display is unable to start correctly.

Beginning with Oracle VM VirtualBox 6.0, the VMSVGA graphics controller is the default controller for guests running Linux operating systems. This issue is more likely to appear if install Oracle Linux 8 on an existing guest that was created on an earlier Oracle VM VirtualBox release. To configure Oracle Linux 8 guests, Oracle recommends that you use Oracle VM VirtualBox 6.0 or later.

(Bug ID 30004543)

Running dnf update glusterfs-* command fails to upgrade previously installed packages

If glusterfs-*.i686 packages exist on an Oracle Linux 8 system which you then upgrade to the next update version, running the dnf update glusterfs* command later fails to upgrade GlusterFS packages.

As a workaround, first remove the glusterfs-*.i686 packages from the system, and then run the dnf update glusterfs* command.

(Bug ID 30279840)

Updating libss package might fail if libss-devel package is installed

The libss package might fail to update if the libss-devel package is installed on the system.

This issue persists if UEK R6 is enabled. However, after updating the kernel and enabling UEK R7, the issue is no longer encountered.

To work around this issue, first remove the libss-devel package from the system. Then, install the corresponding version of this package from the Oracle Linux 8 Distro Builder developer repository (ol8_distro_builder).For example, you can run the following command:

sudo dnf --enablerepo=ol8_distro_builder install libss-devel

Note:

The ol8_distro_builder repository is an unsupported developer repository. You should only enable it for this particular installation action, rather than enabling it globally.

(Bug ID 32005190)

Options for configuring disk cache characteristics during VM creation unavailable in web console

Some options for configuring disk cache characteristics in the web console are currently not available when you are creating a new VM or before the installation starts. For disks that are already added to a VM or when adding a new disk to an already running system, these configuration options are available in the web console.

As an alternative solution, and before the installation begins, use the virt-manager CLI or a similar CLI to configure disk cache characteristics for a newly added disk as well for disks that already exist in the VM.

Oracle recommends that you should use Oracle Linux Virtualization Manager for more complex virtualization requirements. For more information, see https://docs.oracle.com/en/virtualization/oracle-linux-virtualization-manager/.

(Bug ID 30301271)

ACPI error messages displayed on Dell EMC PowerEdge Server during boot

During a system boot of an Intel-based Dell EMC PowerEdge Server, error messages similar to the following might be displayed if the Dell Active Power Controller (DAPC) setting is enabled in the BIOS:

kernel: ACPI Error: No handler for Region [SYSI] (0000000061df8ef3) [IPMI] (20190816/evregion-132)
kernel: ACPI Error: Region IPMI (ID=7) has no handler (20190816/exfldio-265)
kernel: ACPI Error: Aborting method \_SB.PMI0._GHL due to previous error (AE_NOT_EXIST) (20190816/psparse-531)
kernel: ACPI Error: Aborting method \_SB.PMI0._PMC due to previous error (AE_NOT_EXIST) (20190816/psparse-531)
kernel: ACPI Error: AE_NOT_EXIST, Evaluating _PMC (20190816/power_meter-743)

To work around this issue, disable the apci_power_meter kernel module as follows:

echo "blacklist acpi_power_meter" >> /etc/modprobe.d/hwmon.conf

After disabling the apci_power_meter kernel module, reboot the system for the change to take effect.

For environments that do not require the DAPC feature, as an alternative workaround, you can disable the DAPC BIOS setting.

(Bug ID 32105233)

Oracle Linux 8 does not recognize SAS controllers on older Oracle Sun hardware

The Oracle Linux 8 installer does not recognize some Serial Attached SCSI (SAS) controllers that are found in older Oracle Sun server models. If you attempt to install Oracle Linux 8 on these server models, the installer does not recognize the local disk and the installation fails. Examples of these server models include, but are not limited to, the following: Oracle Sun Fire X4170 M2 Server, Oracle Sun Fire X4170 M3 Server, Oracle Sun OVCA X3-2 Server, and the Oracle Sun X4-2 Server.

The following SAS controllers are removed from the mpt2sas driver in RHCK:

SAS2004, PCI ID 0x1000:0x0070
SAS2008, PCI ID 0x1000:0x0072
SAS2108_1, PCI ID 0x1000:0x0074
SAS2108_2, PCI ID 0x1000:0x0076
SAS2108_3, PCI ID 0x1000:0x0077
SAS2116_1, PCI ID 0x1000:0x0064
SAS2116_2, PCI ID 0x1000:0x0065
SSS6200, PCI ID 0x1000:0x007E

The following SAS controllers are removed from the megaraid_sas driver in RHCK:

Dell PERC5, PCI ID 0x1028:0x15
SAS1078R, PCI ID 0x1000:0x60
SAS1078DE, PCI ID 0x1000:0x7C
SAS1064R, PCI ID 0x1000:0x411
VERDE_ZCR, PCI ID 0x1000:0x413
SAS1078GEN2, PCI ID 0x1000:0x78
SAS0079GEN2, PCI ID 0x1000:0x79
SAS0073SKINNY, PCI ID 0x1000:0x73
SAS0071SKINNY, PCI ID 0x1000:0x71

The workaround for this issue to use the Unbreakable Enterprise Kernel Release 6 (UEK R6) boot ISO, and then run UEK R6 with Oracle Linux 8, as these controllers are supported in the Unbreakable Enterprise Kernel release.

(Bug ID 29120478)

File System Issues

The following are known file systems issues that have been encountered in this release of Oracle Linux 8.

Btrfs file system not supported on RHCK

The Btrfs file system is removed from RHCK in Oracle Linux 8, which means you cannot create or mount this file system when using this kernel. Also, any Btrfs user space packages that are provided are not supported with RHCK.

Support for the Btrfs file system is enabled in UEK R7 and UEK R6. Starting with Oracle Linux 8.3, during an installation, you have the option to create a Btrfs root file system, as well as select Btrfs as the file system type when formatting devices.

For further details about these changes, see the following documentation:

For information about creating a Btrfs root file system during an installation, see Oracle Linux 8: Installing Oracle Linux.
For information about managing the Btrfs file system, see Oracle Linux 8: Managing Local File Systems.
For the latest information about other enhancements that have been made to Btrfs in UEK R6, see Unbreakable Enterprise Kernel: Release Notes for Unbreakable Enterprise Kernel Release 6 Update 3 (5.4.17-2136).

For information about UEK R7, see Unbreakable Enterprise Kernel: Release Notes for Unbreakable Enterprise Kernel Release 7 (5.15.0-0.30).

OCFS2 file system not supported on RHCK

The OCFS2 file system is removed from RHCK in Oracle Linux 8, which means you cannot create or mount this file system when using this kernel. Also, OCFS2 user space packages that are provided are not supported with RHCK.

Note that support for OCFS2 file systems is enabled in UEK R7 and UEK R6. For the latest information and other enhancements that have been made to OCFS2 in UEK R6, see Unbreakable Enterprise Kernel: Release Notes for Unbreakable Enterprise Kernel Release 6 Update 3 (5.4.17-2136). See also Unbreakable Enterprise Kernel: Release Notes for Unbreakable Enterprise Kernel Release 7 (5.15.0-0.30).

ext4: Frequent or repeated system shutdowns can cause file system corruption

If a system that is using the ext4 file system is repeatedly or frequently shut down, the file system might become corrupted. This issue is difficult to replicate and is therefore considered to be a corner-case issue. The issue exists in the upstream code and proposed patches are currently under review.

(Bug ID 27547113)

Kernel Issues

The following are known kernel issues that have been encountered in this release of Oracle Linux 8.

Default kernel for KVM guest snapshots might not be interchangeable

Creating a KVM guest snapshot that uses one kernel and then reverting the snapshot to switch to another kernel might cause various virsh commands to fail, as shown in the following examples. The switch can be from RHCK to UEK or vice-versa.

sudo virsh start myGuest --console

error: Failed to start domain myGuest
error: operation failed: guest CPU doesn't match specification: missing
features: md-clear

Running the following command produces similar errors:

sudo virsh snapshot-revert myGuest mySnapshot1 --force

error: operation failed: guest CPU doesn't match specification: missing
features: ibpb,amd-ssbd

However, this issue is not encountered if you are interchanging kernel versions from the same vendor.

To avoid this issue, ensure that you guest snapshots use the same kernel on which they were initially created.

(Bug ID 30561489)

KVM guests boot with "amd64_edac_mod: Unknown symbol" errors on AMD 64-bit platforms

The following errors might be displayed repeatedly when KVM guests are booting on 64-bit AMD hosts:

[   12.474069] amd64_edac_mod: Unknown symbol amd_register_ecc_decoder (err [ 120)
[   12.474083] amd64_edac_mod: Unknown symbol amd_report_gart_errors (err 0)
[   12.852250] amd64_edac_mod: Unknown symbol amd_unregister_ecc_decoder (err 0)
[   12.852297] amd64_edac_mod: Unknown symbol amd_register_ecc_decoder (err 0)
.
.
.

These errors occur because the module code for the kernel erroneously returns -EEXIST for modules that failed to load and are in the process of being removed from the module list. The amd64_edac_mod module will not be loaded in a VM. These errors can be ignored, as they do not impact functionality in any way.

This issue occurs on Oracle Linux 8 hosts that are running RHCK only and is not encountered on UEK R6 hosts.

(Bug ID 29853602)

Output of modinfo command does not show Retpoline support

A bug in the Oracle Linux 8 code causes Retropline support to not be displayed in the output of the modinfo -F retpoline command, even though the CONFIG_RETPOLINE flag is set to Y, for example:

sudo modinfo -F retpoline
/usr/lib/modules/4.18.0-80.el8.x86_64/kernel/sound/usb/usx2y/snd-usb-us122l.ko
.xz

The CONFIG_RETPOLINE=Y flag is still required to add and display Retpoline support. If the parameter is enabled, the kernel builds with a retpoline capable compiler.

To confirm that the CONFIG_RETPOLINE flag is enabled, search for the parameter in the kernel's config-kernel configuration file, for example:

cat /boot/config-5.4.17-2011.7.4.el8uek.x86_64 | grep RETPOLINE.

CONFIG_RETPOLINE=y

(Bug ID 29894295)

Kdump might fail on some AMD hardware

Kdump might fail on some AMD hardware that is running the current Oracle Linux release. Impacted hardware includes the AMD EPYC CPU servers.

To work around this issue, modify the /etc/sysconfig/kdump configuration file and remove the iommu=off command-line option from the KDUMP_COMMANDLINE_APPEND variable. Restart the kdump service for the changes to take effect.

(Bug ID 31274238, 34034614, 34211826)

Limitations of the LVM dm-writecache caching method

The new LVM dm-writecache caching method has certain limitations that do not exist with the dm-cache method, including the following:

Cannot attach or detach dm-writecache when a logical volume is active.
Cannot take a snapshot of a logical volume when the logical volume is using dm-writecache.
Must use a dm-writecache block size that matches the existing file system block size when attaching dm-writecache to an inactive logical volume.
Cannot resize a logical volume when dm-writecache is attached to the volume.
Cannot use pvmove commands on devices that are used with dm-writecache.
Cannot use logical volumes with dm-writecache when using thin pools or the virtual data optimizer (VDO).

For more information about the dm-writecache caching method, see the File Systems and Storage features section of Oracle Linux 8: Release Notes for Oracle Linux 8.2. See also the lvmcache(7) manual page.

aarch64 only: bcache error on UEK R6 during subsequent attempt to register cache device

An attempt to subsequently register a cache set after removing it fails with the following error:

echo "CACHE_DEV" > /sys/fs/bcache/register
        echo: write error: Invalid argument

The following error message is displayed in the dmesg output:

bcache: register_bcache() error /dev/CACHE_DEV: Not a bcache superblock

This issue is related to 64KB page size (CONFIG_ARM64_64K_PAGES=y), which bcache currently does not support.

Because the superblock for the cache device becomes corrupted during this process, it is not possible to re-register the device.

To work around this issue, reinstall the bcache-tools package and then create a new bcache configuration.

As an alternative, avoid using a bcache configuration on Oracle Linux 8 (aarch64) systems.

(Bug ID 30210051)

tracepath6 does not parse destination IPv6 address correctly

Running the tracepath6 command fails to parse the destination IPv6 address correctly. Consequently, the tool traces a route to the wrong host.

To work around this issue, use a tool with similar capabilities to the tracepath6 command.

(Bug ID 29540588)

Running nohup prevents ssh command from executing

On an Oracle Linux 8 system, running the nohup command such as given in the following example might cause ssh command issues.

/usr/bin/nohup ./myscript > nohup.out &

If you attempt to remotely connect to that same system by using the ssh command, the command hangs.

To work around this issue, modify the nohup command syntax as follows:

/usr/bin/nohup ./myscript > nohup.out 2>&1 &

(Bug ID 30287091)

Restarting firewalld service results in SSH connection timeout

Restarting the firewalld service leads to an SSH connection timeout on the terminal from which the service was started. Note that other SSH terminals remain connected.

(Bug ID 29478124)

Error: "mcelog service does not support this processor"

An error indicating that the mcelog service does not support the processor can appear in the system log on systems with AMD processors, such as some Oracle Server hardware. The message might be displayed as follows:

mcelog: ERROR: AMD Processor family
23: mcelog does not support this processor.  Please use the edac_mce_amd
module instead.

The mcelog daemon is a service that is used on x86_64 platforms to log and handle hardware error messaging. On AMD systems, the edac_mce_amd kernel module handles machine exception logging. Therefore, AMD systems do not require the mcelog daemon. This error should be downgraded to a warning.

(Bug ID 29501190)

Power button defaults to ACPI Suspend mode

By default, the Oracle Linux 8 graphical user interface (GUI) console mode treats the hardware power button as the equivalent of the ACPI "Sleep" button, which puts the system into low-power sleep mode. This behavior is specific to the GNOME desktop environment.

In previous Oracle Linux releases, the hardware power button initiated a system shutdown. To ensure that Oracle Linux 8 behaves the same way, do the following:

Create a file named /etc/dconf/db/local.d/01-shutdown-button-action with following content:
```
[org/gnome/settings-daemon/plugins/power]
power-button-action='interactive'
```
Create a file named /etc/dconf/db/local.d/locks/01-power with the following content:
```
/org/gnome/settings-daemon/plugins/power/power-button-action
```
Run the following command:
```
sudo dconf update
```
Log out of the desktop environment and then log back in for the new settings to take effect.

(Bug ID 25597898)

Podman Issues

For information about known issues with the Podman container management tool in Oracle Linux 8, refer to Known Issues chapter of the Oracle Linux: Podman User's Guide.

TLS 1.3 not supported for NSS in FIPS Mode

TLS 1.3 is enabled by default in Oracle Linux 8. Applications that are built with NSS do not support connections that require TLS 1.3 in FIPS mode. To make such connections work, disable FIPS mode or use TLS 1.2.

SAN boot may fail when using an Emulex 32GB FC Adapter

Booting from a multipath attached LUN when using the Emulex 32GB FC Adapter may fail. Note that booting from single path attached LUNs are not affected by this issue. To work around the issue, add the rd.multipath=1 rd.driver.pre=lpfc boot parameter.

For systems with multiple boot disks, such as a local boot disk and a SAN boot disk, the rd.driver.pre=lpfc option does not guarantee that the SAN attached storage is discovered first. To avoid booting from devices other than the SAN, such as from a local disk, additional module_blacklist=disk_driver boot options are required; for example, rd.driver.blacklist=megaraid_sas module_blacklist=megaraid_sas.

(Bug ID 31898488)