58장. Installation and Booting

If you select the Lithuanian (Lietuvių) langauge on the first screen of the graphical installer and press Continue (Tęsti), the installer crashes and displays a traceback message. To work around this problem, either use a different language, or avoid the graphical installer and use a different approach such as the text mode or a Kickstart installation. (BZ#1527319)

The OpenSCAP Anaconda add-on fails to fully remediate a machine to the specified security policy when the system is installed using a Kickstart file that sets installation display mode to the text-based user interface (TUI) using the text Kickstart command. The problem occurs because packages required for the remediation are not installed.

To work around this problem, you can either use the graphical installer or add packages required by the security policy to the %packages section of the Kickstart file manually. (BZ#1547609)

The grub2-mkimage command may fail on UEFI systems with the following error message:

error: cannot open `/usr/lib/grub/x86_64-efi/moddep.lst': No such file or directory.

This error is caused by a the package grub2-efi-x64-modules package missing from the system. The package is missing due to a known issue where it is not part of the default installation, and it is not marked as a dependency for grub2-tools which provides the grub2-mkimage command.

The error also causes some other tools which depend on it, such as ReaR, to fail.

To work around this problem, install the grub2-efi-x64-modules, either manually using Yum, or by adding it to the Kickstart file used for installing the system. (BZ#1512493)

A known issue related to the fix for the Meltdown vulnerability causes a kernel panic with a NULL pointer dereference during the installation of Red Hat Enterprise Linux 7.5 on HPE BL920s Gen2 (Superdome 2) systems. When the problem appears, the following error message is displayed:

WARNING: CPU: 576 PID: 3924 at kernel/workqueue.c:1518__queue_delayed_work+0x184/0x1a0

Then the system reboots, or enters an otherwise faulty state.

There are multiple possible workarounds for this problem:

In Red Hat Enterprise Linux 6, the READONLY=yes option in the /etc/sysconfig/readonly-root file was used to configure a read-only system partition. In Red Hat Enterprise Linux 7, the option is no longer sufficient, because systemd uses a new approach to mounting the system partition.

To configure a read-only system in Red Hat Enterprise Linux 7: