- openEuler Fails to Start After It Is Installed to the Second Disk
- openEuler Enters Emergency Mode After It Is Started
- openEuler Fails to Be Reinstalled When an Unactivated Logical Volume Group Exists
- An Exception Occurs During the Selection of the Installation Source
- Kdump Service Fails to Be Enabled
- Fails to Select Only One Disk for Reinstallation When openEuler Is Installed on a Logical Volume Consisting of Multiple Disks
- openEuler Fails to Be Installed on an x86 PM in UEFI Mode due to Secure Boot Option Setting
- pmie_check Is Reported in the messages Log During openEuler Installation
- Installation Fails when a User Selects Two Disks with OS Installed and Customizes Partitioning
- vmcore Fails to Be Generated by Kdump on the PM with LSI MegaRAID Card Installed
openEuler Fails to Start After It Is Installed to the Second Disk
The OS is installed on the second disk sdb during the installation, causing startup failure.
When openEuler is installed to the second disk, MBR and GRUB are installed to the second disk sdb by default. The following two situations may occur:
- openEuler installed on the first disk is loaded and started if it is complete.
- openEuler installed on the first disk fails to be started from hard disks if it is incomplete.
The preceding two situations occur because the first disk sda is booted by default to start openEuler in the BIOS window. If openEuler is not installed on the sda disk, system restart fails.
This problem can be solved using either of the following two methods:
- During the openEuler installation, select the first disk or both disks, and install the boot loader on the first disk sda.
- After installing openEuler, restart it by modifying the boot option in the BIOS window.
openEuler Enters Emergency Mode After It Is Started
openEuler enters emergency mode after it is powered on.
Damaged OS files result in disk mounting failure, or overpressured I/O results in disk mounting timeout (threshold: 90s).
An unexpected system power-off and low I/O performance of disks may also cause the problem.
Log in to openEuler as the root user.
Check and restore files by using the file system check (fsck) tool, and restart openEuler.
The fsck tool checks and maintains inconsistent file systems. If the system is powered off or a disk is faulty, run the fsck command to check file systems. Run the fsck.ext3 -h and fsck.ext4 -h commands to view the usage method of the fsck tool.
If you want to disable the timeout mechanism of disk mounting, add x-systemd.device-timeout=0 to the etc/fstab file. For example:
# # /etc/fstab # Created by anaconda on Mon Sep 14 17:25:48 2015 # # Accessible filesystems, by reference, are maintained under '/dev/disk' # See man pages fstab(5), findfs(8), mount(8) and/or blkid(8) for more info # /dev/mapper/openEuler-root / ext4 defaults,x-systemd.device-timeout=0 0 0 UUID=afcc811f-4b20-42fc-9d31-7307a8cfe0df /boot ext4 defaults,x-systemd.device-timeout=0 0 0 /dev/mapper/openEuler-home /home ext4 defaults 0 0 /dev/mapper/openEuler-swap swap swap defaults 0 0
openEuler Fails to Be Reinstalled When an Unactivated Logical Volume Group Exists
After a disk fails, openEuler fails to be reinstalled because a logical volume group that cannot be activated exists in openEuler.
During the installation of openEuler, a logical volume group cannot be activated.
Before reinstalling openEuler, restore the abnormal logical volume group to the normal status or clear it. For example:
Restore the logical volume group.
Run the following command to clear the active status of the abnormal logical volume group to ensure that the error message “Can't open /dev/sdc exclusively mounted filesystem” is not displayed:
vgchange -a n testvg32947
Run the following command to recreate a physical volume based on the backup file:
pvcreate --uuid JT7zlL-K5G4-izjB-3i5L-e94f-7yuX-rhkLjL --restorefile /etc/lvm/backup/testvg32947 /dev/sdc
Run the following command to restore the logical volume group information:
Run the following command to reactivate the logical volume group:
vgchange -ay testvg32947
Run the following commands to clear the logical volume group:
vgchange -a n testvg32947 vgremove -y testvg32947
An Exception Occurs During the Selection of the Installation Source
After the installation source is selected, the message “Error checking software selection” is displayed.
This is because the software package dependency in the installation source is abnormal.
Check whether the installation source is abnormal. Use the new installation source.
Kdump Service Fails to Be Enabled
Run the systemctl status kdump command. The following information is displayed, indicating that no memory is reserved.
The kdump service requires the system to reserve memory for running the kdump kernel. However, the system does not reserve memory for the kdump service. As a result, the kdump service cannot be started.
For the scenario where the OS has been installed
Add crashkernel=1024M,high to /boot/efi/EFI/openEuler/grub.cfg.
Restart the system for configuration to take effect.
Run the following command to check the kdump status:
systemctl status kdump
If the following information is displayed, the kdump status is active, indicating that the kdump service is enabled. No further action is required.
The following table describes the parameters of the memory reserved for the kdump kernel.
Table 1 crashkernel parameters
Fails to Select Only One Disk for Reinstallation When openEuler Is Installed on a Logical Volume Consisting of Multiple Disks
If openEuler is installed on a logical volume consisting of multiple disks, an error message will be displayed as shown in Figure 1 when you attempt to select one of the disks for reinstallation.
The previous logical volume contains multiple disks. If you select one of the disks for reinstallation, the logical volume will be damaged.
The logical volume formed by multiple disks is equivalent to a volume group. Therefore, you only need to delete the corresponding volume group.
Press Ctrl+Alt+F2 to switch to the CLI and run the following command to find the volume group:
Run the following command to delete the volume group:
Run the following command to restart the installation program for the modification to take effect:
systemctl restart anaconda
You can also press Ctrl+Alt+F6 to return to the GUI and click Refresh in the lower right corner to refresh the storage configuration.
openEuler Fails to Be Installed on an x86 PM in UEFI Mode due to Secure Boot Option Setting
During the installation of openEuler on an x86 PM in UEFI mode, the system stays at the “No bootable device” page and the installation cannot continue because secure boot is set to enabled (by default, it is set to disabled), as shown in Figure 2.
Figure 2 Dialog box showing “No bootable device”
After secure boot is set to enabled, the mainboard verifies the boot program and OS. If the boot program and OS are not signed using the corresponding private key, they cannot pass the authentication of the built-in public key on the mainboard.
Access the BIOS, set secure boot to disabled, and reinstall the openEuler.
During the system startup, press F11 and enter the password Admin@9000 to access the BIOS.
Choose Administer Secure Boot.
Set Enforce Secure Boot to Disabled.
After Enforce Secure Boot is set to Disabled, save the settings and exit. Then, reinstall the system.
pmie_check Is Reported in the messages Log During openEuler Installation
During the OS installation, if you click Server > Performance tool, PCP is installed. After the OS is installed and restarted, an error “pmie_check failed in /usr/share/pcp/lib/pmie” is displayed in the /var/log/messages log.
anaconda does not support the installation of SELinux policy module in the chroot environment. During the pcp-selinux installation, the postin script fails to execute the PCP-related SELinux policy module. As a result, an error is reported after the OS is restarted.
After the OS is installed and restarted, perform either of the following two operations:
Install SElinux policy module pcpupstream.
/usr/libexec/pcp/bin/selinux-setup /var/lib/pcp/selinux install "pcpupstream"
sudo dnf reinstall pcp-selinux
Installation Fails when a User Selects Two Disks with OS Installed and Customizes Partitioning
During the OS installation, the OS has been installed on two disks. In this case, if you select one disk for custom partitioning, and click Cancel to perform custom partitioning on the other disk, the installation fails.
A user selects a disk for partitioning twice. After the user clicks Cancel and then selects the other disk, the disk information is incorrect. As a result, the installation fails.
Select the target disk for custom partitioning. Do not frequently cancel the operation. If you have to cancel and select another disk, you are advised to reinstall the OS.
Learn More About the Issue at:
vmcore Fails to Be Generated by Kdump on the PM with LSI MegaRAID Card Installed
After the Kdump service is deployed, kernel breaks down due to the manual execution of the echo c > /proc/sysrq-trigger command or kernel fault. When Kdump enables second kernel, an error “BRCM Debug mfi stat 0x2d, data len requested/completed 0x200/0x0” is reported in the MegaRAID driver, as shown in the following figure. As a result, vmcore fails to be generated.
The reset_devices parameter is configured by default and is enabled during second kernel startup, making MegaRAID driver or disk faulty. An error is reported when the vmcore file is dumped ana accesses the MegaRAID card. As a result, vmcore fails to be generated.
Delete the reset_devices parameter in the etc/sysconfig/kdump file on a PM, as shown in the following figure. Therefore, the I/O request will be responded when the MegaRAID driver resets the device during the second kernel startup, and vmcore will be successfully generated.