System Administration Issues

This section describes the system administration issues in Oracle Solaris 11.

System Panics When Multiple USB Keyboards Are Unplugged at the Same Time (7076162)

If you unplug more than one USB keyboard at the same time, a system panic might occur.

Workaround: When you want to unplug more than one USB keyboard from your system, make sure you unplug them one by one.

unconfigure Operation of an IB HCA or the RDSv3 IB Client Fails (7098279)

The unconfigure operation of an InfiniBand host channel adapter (IB HCA) or the RDSv3 IB client fails because the RDSv3 driver does not support dynamic reconfiguration.

The following error message is displayed for RDSv3 DR:

# cfgadm -c unconfigure ib::rdsv3,0
This operation will suspend activity on the IB device
Continue (yes/no)? yes
cfgadm: Hardware specific failure:
unconfigure operation failed ap_id: /devices/ib:fabric::rdsv3,0

The following error message is displayed for HCA DR:

# cfgadm -c unconfigure PCI-EM0
cfgadm: Component system is busy, try again: unconfigure failed

Workaround: Remove the RDSv3 driver and reboot the system.

# pkg uninstall system/io/infiniband/reliable-datagram-sockets-v3

# reboot

Only the First GRUB menu.lst Entry of the Source Boot Environment Is Created When You Create a New Boot Environment (7093604)

When a new boot environment is created, it has only one entry in the GRUB menu.lst file regardless of how many entries the source boot environment has. The source boot environment's first menu.lst entry is used to create the entry for the new boot environment. All other entries are ignored.

This issue occurs when you create the new boot environment either through the beadm command or the pkg update command. Other entries for the new boot environment will not be available.

Workaround: Edit the /rpool/boot/grub/menu.lst file and copy the desired entries from the original boot environment. Replace the boot environment name in the source entries with the name of the target boot environment.

JRE Fails to Detect the System Default Time Zone Setting (7092679)

The Java Runtime Environment (JRE) might fail to detect the system default time zone setting and fall back to the standard time-based custom time zone, such as GMT-08:00 for US/Pacific. As a result, the local time is always in the standard time, and the custom time zone does not support any standard or daylight time transitions.

Workaround: Choose one of the following workarounds:

• Set the TZ environment variable to the appropriate time zone name, such as TZ=US/Pacific. Refer to the environ(5) man page for more details.

• Change the environment/TZ property of the svc:/system/environment:init SMF service instance to the appropriate time zone name and then refresh the service.

  # svccfg -s svc:/system/environment:init setprop environment/TZ = timezone-name
  # svcadm refresh svc:/system/environment:init 

• Specify the user.timezone property value with the appropriate time zone name when starting Java, such as java -Duser.timezone=US/Pacific.

man Command -f Option Is Equivalent to the -k Option (7093992)

Starting with this release, the man command with the -f option is equivalent to the -k keyword option. The man -f command does not display the information as documented in the man page of the man command.

Workaround: None.

zoneadm attach Command Might Hang With the -a and -n Options (7089182)

In this release, the hyphen (-) symbol cannot be used to specify the stdin argument with the -n or -a options of the zoneadm attach command.

Workaround: Do not use the stdin argument.

svccfg validate Command Fails on a Split Manifest (7054816)

With proper multiple manifest support delivered for SMF decorations, the svccfg validate command should no longer warn about missing required property groups. However, the svccfg validate command fails on a split manifest.

This issue occurs when the instance delivering the manifest does not validate correctly without service methods which are defined at the service level in a separate manifest. The following error message might be displayed:

$ svccfg validate /lib/svc/manifest/system/console-login-vts.xml
Required property group missing: FMRI="svc:/system/console-login:vt6"; 
Name of missing property group="start"; Type of missing property group="method"; 
Template source="svc:/system/svc/restarter:default"; 
pg_pattern name="start"; pg_pattern type="method"

Required property group missing: FMRI="svc:/system/console-login:vt6"; 
Name of missing property group="stop"; Type of missing property group="method"; 
Template source="svc:/system/svc/restarter:default"; 
pg_pattern name="stop"; pg_pattern type="method"

Required property group missing: FMRI="svc:/system/console-login:vt5"; 
Name of missing property group="start"; Type of missing property group="method"; 
Template source="svc:/system/svc/restarter:default"; 
pg_pattern name="start"; pg_pattern type="method"

Required property group missing: FMRI="svc:/system/console-login:vt2"; 
Name of missing property group="stop"; Type of missing property group="method"; 
Template source="svc:/system/svc/restarter:default"; 
pg_pattern name="stop"; pg_pattern type="method"

svccfg: Validation failed.

The warning is displayed in the manifest-import service log file rather than on the console. The warning is not fatal and does not prevent the import of the instance. Once the instance is imported, you can successfully validate the instance manually by using the svccfg validate command.

Workaround: Avoid using the svccfg validate command on split manifests.

ilomconfig Service Might Delay System Boot (7100050)

For desktop and laptop systems that do not have a Service Processor using Oracle Integrated Lights Out Manager (ILOM), or using Oracle ILOM older than version 3.0.12, the ilomconfig-interconnect service repeatedly attempts to establish a connection to Oracle ILOM. This repeated attempts might delay the system boot time.

Workaround: Disable the ilomconfig service by using the following command:

# svcadm disable svc:/network/ilomconfig-interconnect

xterm-256color Terminal Type Is Not Supported (7099457)

The Terminal.app terminal emulator included in Apple OS X Lion (Mac OS X Version 10.7) uses the default xterm-256color terminal type, which is not supported in this Oracle Solaris release. Therefore, when you run remote terminal sessions on Oracle Solaris 11 from Apple OS X Lion, the available terminal capabilities are significantly restricted and only very basic command-line interaction is possible.

You might see error messages whenever applications which need to retrieve terminal capabilities from the terminfo database are run. Typical examples of such error messages are as follows:

tput: unknown terminal "xterm-256color"
tabs: xterm-256color: terminfo file not found

Workaround: Configure the Terminal.app terminal emulator to use the xterm-color terminal type. Perform the following steps to set the xterm-color terminal type:

1. When the Terminal.app application is focused, from the menu at the top of the screen, choose Terminal -> Preferences -> Settings -> Advanced.

2. From the Declare Terminal As drop-down list, select xterm-color.

Zone Boot Environment Datasets Should Be Mounted When Zone Is Ready (7091692)

The pkg(1) command will fail if any zone is in the ready state. The following error message is displayed:

pkg: update failed (linked image exception(s)):
Can't initialize child image (zone: zone) at path: zonepath/root

Workaround: Boot or halt the zone by using the zoneadm command. Then, retry the pkg command.

Disconnecting or Powering off a Connected IEEE-1394 Device Causes System Panic (7070912)

Disconnecting or powering off a connected IEEE-1394 device might cause the system to panic. The following error message is displayed:

panic[cpu3]/thread=ffffff0010633c20:
genunix: [ID 335743 kern.notice] 
BAD TRAP: type=e (#pf Page fault) rp=ffffff0010633960 
addr=8 occurred in module "s1394" due to a NULL pointer dereference

Workaround: Do not disconnect or power off the connected IEEE-1394 devices.

When the automount Service Is Disabled, useradd Creates Users With Inaccessible Home Directories (7070558)

If the automount service is disabled, users whose accounts are created by using the useradd(1M) command will be unable to log in. This issue occurs because the home directory specification in the /etc/passwd file does not correspond to the actual path name of the home directory. The following error message is displayed:

No directory!

Workaround: Do not disable the automount service, svc:/system/filesystem/autofs:default.

The automount service must be enabled because existing user accounts including the initial account created at installation time depend on the automount service. If you choose to disable the automount service, manually edit the home directory entry for each user in the /etc/passwd file to specify the actual pathname of the home directory.

SPARC: Using Direct I/O to Remove Multiple PCIe Slots From the Primary Domain on a Multisocket T-Series System Might Cause System Panic at Boot Time (7100859)

Your system might panic at boot time if you use the direct I/O feature to remove multiple PCIe slots from a multisocket SPARC T-series system. You can use the ldm remove-io command to remove multiple PCIe slots. The issue occurs when the PCIe slots that have similar path names are removed with the ldm remove-io command and the system is rebooted.

For more information about the direct I/O feature, see the Oracle VM Server for SPARC 2.1 Administration Guide.

For example, if you remove the /SYS/MB/PCIE5 (pci@500/pci@2/pci@0/pci@0) and /SYS/MB/PCIE4 (pci@400/pci@2/pci@0/pci@0) slots, which have similar path names, the next boot of the Oracle Solaris 11 OS might panic.

When the administrator runs the ldm list-io command, after removing the /SYS/MB/PCIE4 and /SYS/MB/PCIE5 PCIe slots, the following configuration is displayed:

# ldm list-io
IO              PSEUDONYM       DOMAIN
--              ---------       ------
pci@400         pci_0           primary
niu@480         niu_0           primary
pci@500         pci_1           primary
niu@580         niu_1           primary

PCIE                       PSEUDONYM       STATUS  DOMAIN
----                       ---------       ------  ------
pci@400/pci@2/pci@0/pci@8  /SYS/MB/PCIE0   OCC     primary
pci@400/pci@2/pci@0/pci@4  /SYS/MB/PCIE2   OCC     primary
pci@400/pci@2/pci@0/pci@0  /SYS/MB/PCIE4   OCC
pci@400/pci@1/pci@0/pci@8  /SYS/MB/PCIE6   OCC     primary
pci@400/pci@1/pci@0/pci@c  /SYS/MB/PCIE8   OCC     primary
pci@400/pci@2/pci@0/pci@e  /SYS/MB/SASHBA  OCC     primary
pci@400/pci@1/pci@0/pci@4  /SYS/MB/NET0    OCC     primary
pci@500/pci@2/pci@0/pci@a  /SYS/MB/PCIE1   OCC     primary
pci@500/pci@2/pci@0/pci@6  /SYS/MB/PCIE3   OCC     primary
pci@500/pci@2/pci@0/pci@0  /SYS/MB/PCIE5   OCC
pci@500/pci@1/pci@0/pci@6  /SYS/MB/PCIE7   OCC     primary
pci@500/pci@1/pci@0/pci@0  /SYS/MB/PCIE9   OCC     primary
pci@500/pci@1/pci@0/pci@5  /SYS/MB/NET2    OCC     primary

Workaround: Use one of the following workarounds:

• Do not remove all the PCIe slots that have similar paths. Instead, remove only one PCIe slot.

• Insert the PCIe cards into slots that do not have similar paths. Then, use the PCIe slots with the direct I/O feature.

tar Command Fails to Extract Archive From Tapes (7028654)

The tar command might fail to list archive contents or extract the archive from tapes. The following error message might be displayed:

tar: blocksize = 0

Workaround: Choose one of the following workarounds:

• Copy the tape archive to a disk by using the dd command before using the tar command. For example, to update an existing file on the tape archive, you would type the following commands:

  # dd if=/dev/rmt/6c of=disk.tar bs=10240
  # tar uvf disk.tar file
  # dd if=disk.tar of=/dev/rmt/6c bs=10240

• Pipe the archive from the tape to tar.

  # dd if=/dev/rmt/6c bs=10240 | tar tvf -

SPARC: System Fails to Boot When SP Is in Degraded Mode (6983290)

On T3 and T4 systems, when the service processor (SP) is in degraded mode, the host system might fail to reboot.

Workaround: None.