Chapter 2 Known Issues

Table of Contents

2.1 Configuring Encryption and /boot During Installation
2.2 Network Installation
2.3 Installation Fails for ext3 / and /boot
2.4 grubby Sets Incorrect Saved Entry
2.5 Per-CPU Allocation Fails when Loading kvm_intel Module with UEK R3
2.6 avahi-daemon Fails to Start with UEK R3
2.7 Crash Kernel auto Setting
2.8 systemd Fails to Load the autofs4 and ipv6 Modules with UEK R3
2.9 firewalld Does Not Currently Support IPv6 NAT Under UEK R3
2.10 Oracle ASM Fails to Initialize with SELinux in Enforcing Mode
2.11 systemctl Does Not Support Some Service Actions
2.12 btrfs-convert Fails
2.13 Upgrading from Oracle Linux 6 Update 5
2.14 Oracle Linux 7 Guests on Oracle VM 2
2.15 Hebrew LaTeX Fonts
2.16 sosreport Reports Many SELinux Warnings
2.17 Using NFS v4 with an lxc-oracle Container Fails
2.18 Enabling and Disabling NFS
2.19 Network Teaming
2.20 Network Connection Icon Reports Incorrect State for Interfaces
2.21 net_prio Control Group
2.22 XFS File Attributes Cause a Panic or Reboot
2.23 Automatic Bug Reporting Tool
2.24 InfiniBand Issues

This chapter describes the known issues for the initial release for Oracle Linux 7.

2.1 Configuring Encryption and /boot During Installation

During installation, if you select Encrypt my data on the Installation Destination screen and then perform manual partitioning, the Encrypt check box is not shown as selected on the Manual Partitioning screen. This check box refers to encryption that you can configure on a file system type that supports encryption or on an LVM logical volume that contains the file system. If you click Modify, the Encrypt check box on the Configure Volume screen is shown as selected for the volume, meaning that the encryption will be applied at the level of the underlying block device.

For LVM, selecting Encrypt my data encrypts the LVM physical volume and all the logical volumes that it contains. If you do not select Encrypt my data, you can encrypt the logical volume by selecting the Encrypt check box on the Manual Partitioning screen or encrypt the physical volume by selecting the Encrypt check box on the Configure Volume screen.

For btrfs, encryption can only be applied to the block device that contains the file system, including its subvolumes. For example, enabling encryption for the /home subvolume of a btrfs root file system implicitly enables encryption for the root file system itself. You can only select the Encrypt check box on the Configure Volume screen. As btrfs does not support encryption at the file-system level, you cannot select the Encrypt check box on the Manual Partitioning screen for a btrfs file system.

Do not select the Encrypt check box or a BTRFS, LVM, or LVM Thin Provisioning device type for /boot. The /boot file system must be configured on a standard partition and should be of type ext4 or XFS.

When entering a password in the Disk Encryption Passphrase dialog, press Tab to move between the entry fields. You cannot use the mouse to select the fields.

2.2 Network Installation

Attempting to perform a network installation without configuring a network interface to use DHCP to obtain its IP settings or with static IP settings results in the error Error in Installation Source.

For example, if you use a feature such as a remote console or Lights-out management to access a boot ISO, the network configuration of the embedded server manager might not be available when you select the installation location. The workaround is to use the graphical installer to configure the network settings manually before configuring the installation location. (Bug ID 19047736)

2.3 Installation Fails for ext3 / and /boot

Rebooting following installation fails if you format / and /boot as ext3 file systems. The system displays messages similar to the following before starting the emergency shell:

systemd[1]: Timed out waiting for device ...
systemd[1]: Dependency failed for /boot. 
systemd[1]: Dependency failed for Local file ...

systemd[1]: Dependency failed for Mark the need to relabel after reboot. 
systemd[1]: Dependency failed for Relabel all filesystems, if necessary. 
systemd[1]: Dependency failed for File System Check on ...

[DEPEND] Dependency failed for Local File Systems. 
[DEPEND] Dependency failed for Mark the need to relabel after reboot 
[DEPEND] Dependency failed for Relabel all filesystems, if necessary. 

The workaround is to disable SELinux during installation by specifying selinux=0 on the installation boot command line or by specifying selinux --disabled in the Kickstart file. (Bug ID 19171480)

2.4 grubby Sets Incorrect Saved Entry

If grubby is used to remove a kernel menu entry from the GRUB 2 configuration, the value of the default entry in /etc/grub2/grub.cfg is incorrect. The workaround is to set the value of GRUB_DEFAULT in /etc/default/grub to the correct entry and use grub2-mkconfig to recreate /etc/grub2/grub.cfg, or use yum or rpm to remove the kernel packages. (Bug ID 19192278)

2.5 Per-CPU Allocation Fails when Loading kvm_intel Module with UEK R3

Per-CPU allocation fails when the kvm_intel module is loaded with UEK R3. Messages such as the following are logged:

kvm_intel: Could not allocate 48 bytes percpu data 
PERCPU: limit reached, disable warning 

There is no current workaround for UEK R3. (Bug ID 18459498)

2.6 avahi-daemon Fails to Start with UEK R3

The avahi-daemon fails to start with UEK R3 and messages such as the following are logged:

WARNING: No NSS support for mDNS detected, consider installing nss-mdns! 
Unit avahi-daemon.service entered failed state.

The workaround is to comment out the disallow-other-stacks entry in /etc/avahi/avahi-daemon.conf as shown here:

#disallow-other-stacks=yes

(Bug ID 18459758)

2.7 Crash Kernel auto Setting

If you enable the crashkernel=auto kernel parameter for UEK R3 to simplify Kdump configuration, both dmesg output and /proc/cmdline show crashkernel=NNNM@0M. This is the expected behavior for the implementation, where @0M implies the auto setting. For Xen, crashkernel=auto is supported only for Domain 0.

2.8 systemd Fails to Load the autofs4 and ipv6 Modules with UEK R3

At boot time, systemd fails to load the autofs4 and ipv6 modules and errors such as the following are logged:

systemd[1]: Failed to insert module 'autofs4'
systemd[1]: Failed to insert module 'ipv6'

There is no current workaround for UEK R3. (Bug ID 18470449)

2.9 firewalld Does Not Currently Support IPv6 NAT Under UEK R3

The following error message indicates that IPv6 NAT is not currently supported by firewalld with UEK R3.

ERROR: ipv6 table 'nat' does not exist (or not enough permission to check)

(Bug ID 18504545)

2.10 Oracle ASM Fails to Initialize with SELinux in Enforcing Mode

The oracleasm script fails if SELinux is in Enforcing mode. The suggested workaround is to disable the SELInux policy module for Oracle ASM before running oracleasm:

# semodule -d oracleasm
# semodule -l | grep oracleasm
oracleasm      1.0.0   Disabled

(Bug ID 18513404)

2.11 systemctl Does Not Support Some Service Actions

The systemctl command supports the disable, enable, restart, start, status, and stop actions for services such as o2cb and oracleasm. To perform actions such as configure, invoke the /etc/init.d script for the service directly. (Bug IDs 18527520 and 18528039)

2.12 btrfs-convert Fails

The btrfs-convert command fails when invoked to convert an ext2, ext3, or ext4 file system to btrfs. This error occurs whether the system has been booted with UEK R3 or RHCK. There is currently no known workaround or fix available. (Bug ID 18534417)

2.13 Upgrading from Oracle Linux 6 Update 5

The following issues may be encountered when upgrading:

  • The redhat-upgrade-tool-cli utility requires that you install version 3.2.29-43.0.1 or later of the yum package on the Oracle Linux 6 Update 5 system that you want to upgrade. If you use an earlier version of the yum package, the upgrade tool fails with dependency errors. (Bug ID 18648783)

  • The old RHCK is not removed during the upgrade. (Bug ID 18767222)

  • If you do not run the preupgrade assistant utility preupg, an upgraded system hangs while rebooting with the message starting wait for plymouth boot screen to quit. (Bug ID 18815298)

  • The redhat-upgrade-tool-cli utility does not exit if UEK R3 is not installed. (Bug ID 18900135)

  • The postupgrade scripts fail if a proxy is required to access Oracle Public Yum. (Bug ID 19169163)

  • The libcgroup package in Oracle Linux 7 does not include the cgconfig and cgred control group services. To restore these services on an upgraded system, install the libcgroup-tools package. (Bug ID 19177606)

2.14 Oracle Linux 7 Guests on Oracle VM 2

Oracle Linux 7 guests are not supported on Oracle VM 2.

2.15 Hebrew LaTeX Fonts

Installing the tex-fonts-hebrew package fails unless you first install all texlive* packages. (Bug ID 19059949)

2.16 sosreport Reports Many SELinux Warnings

The sosreport -o selinux -a command reports many SELinux warnings for files that do not belong to RPMs. This is the expected behavior. You can safely ignore these warnings. (Bug ID 18913115)

2.17 Using NFS v4 with an lxc-oracle Container Fails

Attempting to create an lxc-oracle container on a remote file system mounted using NFS v4 fails. In addition, attempting to mount a remote file system using NFS v4 from within an lxc-oracle container also fails. The workaround is to use NFS v3 instead. (Bug ID 16316266)

2.18 Enabling and Disabling NFS

You cannot enable or disable the NFS service by using the systemctl command with nfs.service. Specify nfs-server.service instead, for example:

# systemctl enable nfs-server

(Bug ID 18437212)

2.19 Network Teaming

Network teaming is not currently supported for use with UEK R3. The workaround is to use bonding instead. (Bug ID 19151770)

2.20 Network Connection Icon Reports Incorrect State for Interfaces

The network connection icon might report an active network interface as being disconnected. This behavior is seen for the root user but not for other users. Command-line utilities such as ip link and ifconfig report the correct state. (Bug ID 19060089)

2.21 net_prio Control Group

The Network Priority cgroup subsystem (net_prio) is not currently supported for use with UEK R3. Attempting to use the module with UEK R3 results in error messages such as the following:

modprobe: FATAL: Module netprio_cgroup not found
mount: special device cgroup does not exist.

(Bug ID 18966564)

2.22 XFS File Attributes Cause a Panic or Reboot

Attempting to add many attributes to a file in an XFS file system under UEK R3 can result in a kernel panic or a system reboot. (Bug ID 18504299)

2.23 Automatic Bug Reporting Tool

The daemons and features provided by the Red Hat Automatic Bug Reporting Tool (ABRT) are not supported with Oracle Linux. ABRT packages and associated files, such as libreport, are included in the distribution to satisfy package dependencies, but the features within these packages are not supported. For technical assistance, contact Oracle Support via the My Oracle Support portal or by telephone.

2.24 InfiniBand Issues

The following issues might be encountered with InfiniBand devices:

  • You might see the following warning messages if you use the ibportstate disable command to disable a switch port:

    ibwarn: [2696] _do_madrpc: recv failed: Connection timed out
    ibwarn: [2696] mad_rpc: _do_madrpc failed; dport (Lid 38)
    ibportstate: iberror: failed: smp set portinfo failed

    You can safely ignore these warnings. (Bug ID 16248314)

  • To configure Internet Protocol over InfiniBand (IPoIB):

    1. Edit the /etc/sysconfig/network-scripts/ifcfg-ibN configuration file, where N is the number of the interface. The following example shows the configuration for the interface ib0:

      DEVICE=ib0
      TYPE=InfiniBand
      ONBOOT=yes 
      DHCP_HOSTNAME="myhost.mydom.com"
      BOOTPROTO=static
      IPADDR=192.168.100.1
      NETMASK=255.255.255.0
      IPV4_FAILURE_FATAL=yes
      IPV6INIT=no 
      CONNECTED_MODE=no
      NAME=ib0 
    2. Stop the NetworkManager service:

      # systemctl stop NetworkManager
    3. Start the RDMA service:

      # systemctl start rdma
    4. Bring up the interface:

      # ifup ibN

      (Bug ID 19150870)

  • The IPoIB driver supports the use of either connected mode or datagram mode with an interface, where datagram mode is the default mode. Changing the mode of an InfiniBand interface by echoing either connected or datagram to /sys/class/net/ibN/mode is not supported. It is also not possible to change the mode of an InfiniBand interface while it is enabled.

    To change the IPoIB mode of an InfiniBand interface:

    1. Edit the /etc/sysconfig/network-scripts/ifcfg-ibN configuration file, where N is the number of the interface:

      • To configure connected mode, specify CONNECTED_MODE=yes in the file.

      • To configure datagram mode, either specify CONNECTED_MODE=no in the file or do not specify this setting at all (datagram mode is enabled by default).

      Note

      Before saving your changes, make sure that you have not specified more than one setting for CONNECTED_MODE in the file.

    2. To enable the specified mode on the interface, use the following commands to take down the interface and bring it back up:

      # ifdown ibN
      # ifup ibN

    (Bug ID 17479833)

  • When the rds_ib_srq parameter for the rds_rdma module is enabled and the module is in use (for example, when running the rds-stress tool), restarting the rdma service (which reloads the rds_rdma module) generates error messages visible in dmesg or /var/log/messages. (Bug ID 19010606)