| Solaris |
|
|
This section provides general information such as upgrade issues and behavior changes in the OpenSolaris 2009.06 OS.
The OpenSolaris Update Manager automatically detects when new packages are available for your system, including when the Update Manager itself needs to be updated. In this case, Update Manager will prompt you to open a terminal window and execute the following command:
$ pfexec pkg install SUNWipkg
After executing this command, you must close the Update Manager and then restart it via the System-> Administration menu to complete the update.
General instructions for updating from the OpenSolaris 2008.05 release to later versions of OpenSolaris are as follows:
$ BUILD=`uname -v | sed -e "s/snv_//" -e "s/[a-z]//"`
$ pfexec pkg refresh
$ pfexec pkg install entire@0.5.11-0.${BUILD}
$ pfexec pkg install SUNWipkg@0.5.11-0.${BUILD}
$ pfexec pkg install SUNWinstall-libs
$ pfexec pkg image-update
At this point, you can boot into the updated BE using reboot(1M) or init(1M) as usual.
OpenSolaris supports installation through its Live CD on systems or virtual machines that contain
at least 512 MB of RAM. The following limitations exist on systems with low memory.While basic desktop functions and installation functions work well, attempting to use the Live CD
as a full-blown user environment might cause issues. However, an installed system functions normally with 512 MB of RAM.The Live CD supports booting the OpenSolaris kernel in either 32-bit mode or 64-bit mode depending
on the capabilities of the underlying hardware. The OS will boot in the 64-bit mode only if the system has at least 1000 MB of RAM. Systems with less than this amount will boot in the 32-bit mode. This limitation does not apply to the installed system.It is not possible to start the 'Read the Release Notes' link from within the GUI Installer
on systems with 512 MB of RAM. Attempting to open this link causes the system to freeze while loading Firefox. Therefore users are recommended to read the Release Notes prior to launching the GUI Installer.On systems with less than 700 MB of RAM the installer will create a dedicated swap slice instead
of a ZFS volume to use for system swap.For more information about the enhancements that were made to the OS from the OpenSolaris 2008.11 release onwards to
support installation on systems with low memory, see href="http://defect.opensolaris.org/bz/show_bug.cgi?id=3597"># 3597 Unable to install OpenSolaris with 512 MB Base Memory.This section describes known issues that are related to the OpenSolaris 2009.06 documentation.
Some links are incorrect in the Getting Started With OpenSolaris 2009.06 Guide.
The following are the correct and updated links:
The following is the correct and updated link:
This section describes installation issues that apply to the OpenSolaris 2009.06 OS.
If, during installation, no root password is entered and no user account is created, you will not be able to log in to the installed system.
Choose one of the following workarounds:
Enter a root password as follows:
During boot, type “e” in the grub menu twice, in order to edit the boot entry.
Add -s to the end of the boot entry, and boot with this line.
Log in as root and use the passwd command to enter a password.
Restart the system. Boot normally, and log in with the new password.
Set root so that it does not require a password as follows:
During boot, type “e” in the grub menu twice, in order to edit the boot entry.
Add -s to the end of the boot entry, and boot with this line.
Log in to the system as root.
Edit the /etc/X11/gdm/custom.conf file so that the value of PasswordRequired is false.
When creating a Solaris partition, the GUI installer only utilizes the free space at the end of the disk after the last primary partition. However, the sum of the total space available including gaps between partitions is reported on the disk screen. If the user tries to create a Solaris partition which is larger than the amount of free space available at the end of the disk, the size of Solaris partition is adjusted.
A warning message is displayed when the Next button is pressed on the disk screen and the Solaris partition is shrunk. The title of the warning message is as follows:
Adjustments were made to the new partitions
Use the Solaris fdisk(1M) tool or the GParted tool to create a Solaris2 partition. These tools enable the utilization of unallocated space between already defined partitions. Select the newly created Solaris2 partition as the target for the installation on the installer disk screen.
Renaming a boot environment (BE) might fail if the BE has a dependent clone that is currently mounted or otherwise busy.
A typical example of this situation is when the live BE has been successfully upgraded. If you then try to rename the auto-generated and upgraded BE, the renaming fails. The renaming fails because the upgraded BE has been activated and the live BE is now its clone.
Use the following procedure:
Activate the live BE.
Rename the upgraded BE.
Activate the upgraded BE.
For example, if the live BE name is “opensolaris”, and the auto-generated BE name of the upgraded BE is “opensolaris-1,” use the following commands:
# beadm activate opensolaris
# beadm rename opensolaris-1 new_name
# beadm activate new_name
The new fast reboot feature enables you to provide a '-e' option to specify the name of the boot environment (BE) to which you need to fast reboot. Attempting to do a fast reboot and specifying the name of a BE fails in the OpenSolaris 2009.06 release. The following error message is displayed:
# reboot -f -e opensolaris-7
reboot: cannot mount BE opensolaris-7
Specify the full root dataset path to the BE to which you need to fast reboot. For example:
# reboot -f rpool/ROOT/opensolaris-7
If a ZFS pool named rpool is already present on the system because it was manually imported or created by the user during the current boot of the live CD, the install process fails. The last message in the installation log file is as follows:
Root pool rpool exists, we can't proceed with the installation.
This behavior is intentional and ensures that the installer does not inadvertently corrupt the user's data.
The installer recognises cases when rpool is created by the installer but the installation process has failed or was interrupted. In such cases, the installer destroys rpool automatically after it is restarted.
Choose one of the following workarounds:
To preserve data in rpool, reboot the system and do not import the pool. The pool remains invisible to the installer. However, if the pool was created on the target disk then the pool cannot be preserved.
If you do not want to preserve data in rpool, destroy the pool before you start the installer by using the following command:
$ pfexec zpool destroy -f rpool
If a system has a ZFS file system, the Distribution Constructor does not recognise or treat a build area as a ZFS file system in the following cases:
A new sub-directory of the ZFS file system is specified as a build area, but the mountpoint is specified instead of the zpool.
The build area already exists as a ZFS file system but the mountpoint is specified instead of the zpool
For example, in the following zpool:
$ zfs list | grep "disk2_pool/ib/pia"
disk2_pool/ib/pia 1.90G 355G 18K /export/home/ib/pia
When you run the following command:
# distro_const build -p 1 slim_cd.xml
Where the build area in slim_cd.xml is specified as build_area/export/home/ib/pia/build_area, the following message error message is displayed:
Checkpointing is not available
Rerun the build without -p
Specify the zpool in the manifest. For example, build_area/disk2_pool/ib/pia/build_area.
When a new boot environment (BE) is created, the new BE will only have one entry in the GRUB menu.lst file regardless of how many entries the source BE has. The source BE's first menu.lst entry is used to create the entry for the new BE. All other entries are ignored. This issue occurs when you create the new BE either through beadm or pkg image- update.
If the user needs the other entries for the new BE they are not available.
Edit the /rpool/boot/grub/menu.lst file and copy the desired entries from the original BE. Replace the BE name in the source entries with the BE name of the target.
The Distribution Constructor might display the following error message after successfully performing its function:
Unhandled exception in thread started by
Error in sys.excepthook:
Original exception was:
This error message is displayed just before termination and can be safely be ignored.
Use the dumpadm(1M) command to enable crash dumps at boot time:
# dumpadm -y
Alternatively, create a cron job that enables saving crash dumps at any given later time.
The latest version of SUNWinstalladm-tools from http:pkg.opensolaris.org/release contains fixes for the following bug:
#7218 - installadm needs to move service data into SMFIf you update your automated install server with the versions of SUNWinstalladm-tools and SUNWinstall-libs packages that fix #7218, then you will need to convert any services previously created with installadm create-service.
Use the following procedure to convert the services created with installadm create-service.
svcadm enable svc:/system/install/server:default
For more information, see the OpenSolaris Automated Installer Guide.
Automated Installation fails when the Automated Installation manifest points to a target device and the slice 0 on target device is configured with less than 8 GB of disk space. The following error message is displayed:
Auto install failed
The Automated Installation fails because it runs out of space in slice 0 of the target device.
Choose one of the following workarounds:
<ai_target_device> <target_device_name>c0t0d0</target_device_name> <target_device_install_slice_number>4</target_device_install_slice_number> </ai_target_device>
When a custom Automated Installation manifest is used to install a system and there are customizations in the manifest to create slices other than the root slice if those slices already exist on the disk in your system, then the automated installation fails. The following error message is seen in the install_log file:
root@opensolaris:/tmp# tail -20 /tmp/install_log/tmp/ai_combined_manifest.xml is a valid manifest Auto reboot disabled Disk = c8t0d0 found on the system Disk name selected for installation is c8t0d0 no manifest partition information found disk partition info not changed System reports enough physical memory for installation, swap is optional creating slice which already exists failed to modify slice(s) specified in the manifest Auto install failed
Use the format(1M) command to remove the existing slices from your disk. Rerun the automated installation by rebooting your system back into the automated install.
If you type installadm create-service followed immediately by installadm delete-service, the installadm command might hang. This command hangs because the installadm smf service, svc:/system/install/server:default, is not online when the installadm delete-service command is executed.
Wait for the svc:/system/install/server:default service to be online. To determine whether this service is online, type svcs svc:/system/install/server:default and check whether the STATE is listed as online.
PXE boot installation using Automated Installation fails for xVM hvm guests. On the guest console, after displaying the grub entry to boot, the system proceeds to download the kernel and boot archive from the automated install server. The OS banner appears. However, shortly afterwards, the guest cannot connect to the network to download the required installation files and proceed. The installing guest drops to maintenance mode, and prompts for a username and password.
The guest client to be installed by using Automated Installation should be booted with additional kernel options in the menu.lst file which is retrieved from the boot server.
Use any one of the following methods to set the correct menu.lst entries for the client on the automated install server:For individual clients, type the installadm create-client command to specify the extra boot options needed for each installing client. The command syntax is as follows:
installadm create-client -b network-interface=xnf0,xpv-hcp=dhcp -e MAC address of the client -t target image -n install service nameFor example:
# installadm create-client -b network-interface=xnf0,xpv-hcp=dhcp -e 00:16:3e:7f:a9:6f -t /space/images/2009.06_111b -n osol_111bFor all clients of a given automated install service, manually edit the appropriate menu.lst file in /tftpboot to add the network-interface and xpv-hcp boot options in the previous example.
The menu.lst file is named according to the following pattern:
menu.lst.<AI service name>For example, for the automated install service osol, the menu.lst file is /tftpboot/menu.lst.osol and has the following structure:
default=0After adding the required boot options, the menu.lst file has the following structure:
default=0For more information, see the OpenSolaris Automated Installer Guide.
Automated Installation might fail if version strings for packages are specified in the Automated Installation manifest. The following error message might be found in the Automated Installation SMF log at /var/svc/log/application-auto-installer:default.log.
Unable to install SUNW in /a om_perform_install failed with error 114 Auto install failed
List the entire package first and specify version strings for the entire package only. For example:
<ai_install_packages>
<pkg name="entire@0.5.11,5.11-0.111"/>
<pkg name="SUNWcsd"/>
<pkg name="SUNWcs"/>
<pkg name="SUNWj6rtx"/>
<pkg name="SUNWgroff"/>
<pkg name="SUNWgcc"/>
<pkg name="babel_install"/>
<pkg name="xvm"/>
</ai_install_packages>
On disks larger than 1.96 TB, the installer creates VTOC slices which are not aligned on the cylinder boundary. The installed OpenSolaris instance will boot successfully but the format(1M) tool cannot process the VTOC configuration. The tool will report that the disk does not contain a valid label.
Use the Solaris fdisk(1M) and format(1M) tools to create the required VTOC configuration outside the installer. Then install into the existing slice using the Automated Installer.
For more information on how to configure and use the Automated Installer, see the OpenSolaris Automated Installer Guide.
The new Solaris installer installs the Solaris operating system into the Solaris2 partition (0xbf). The installer does not support multiple Solaris partitions on one disk. However, the installer recognizes the Linux swap partition (0x82) correctly and installs successfully if the primary partition is used as the Linux swap partition.
Solaris installation fails when the partition configuration is as following:
The Linux swap primary partition precedes the Solaris2 partition
The Solaris2 partition is not active
Choose one of the following workarounds:
Avoid defining a Linux swap partition as a primary partition on the same disk where the Solaris2 primary partition will be created. For example, create the Linux swap partition as logical volume within extended partition.
Ensure that the Solaris2 partition precedes the Linux swap partition.
Create the Solaris2 partition outside the installer and mark it as active. Then install into that preexisting partition.
Before installation, change the Linux swap partition type to another type, excluding (0xbf) or (0x82). After the installation is complete, change the Linux swap partition type back to (0x82).
If you try to install the OpenSolaris 2009.06 OS on a system that has Fiber Channel disks attached, /usr/bin/gui-install dumps the core.
Disconnect the Fiber Channel disks and restart the installation.
bootadm causes ZFS to hang while installing OpenSolaris 2009.06. No error message is displayed. Install hangs at 99% completion.
Ensure that the Linux swap partition is not a primary partition. Check if the Linux swap partition is a logical partition. To check if it's a logical partition, delete the Linux swap and recreate it as a logical partition. Once this is done the installation is successful.
See also, #8693.
The OpenSolaris 2009.06 OS should not be booted from a disk whose capacity exceeds 232 sectors or 2 TB when the sector size is 512 bytes. This is because of a limitation in the current implementation of GRUB which OpenSolaris uses as the boot loader.
None.
The message displayed when you boot an out-of-sync boot archive advises the user to boot from the GRUB menu "Solaris Failsafe" option . No such option is available on OpenSolaris.
None.
This section describes IPS-related issues that apply to the OpenSolaris 2009.06 OS.
While updating an image to build 111 or beyond, a warning message similar to the following might be displayed:
driver (emlxs) upgrade (removal of minor perm '* 0600 root sys'') failed with return code 252 command run was: /usr/sbin/update_drv -b /mnt -d -m * 0600 root sys' emlxs command output was:
No entry found for driver (emlxs) in file (/mnt/etc/minor_perm).
This bug occurs only when the SUNWemlxs package is manually added to the OpenSolaris 2008.11 OS and the system is updated to the OpenSolaris 2009.06 release. This warning message is displayed for many drivers including emlxs, afe, audio810, audiohd, audioixp, bnx, isa, ixgbe, nge, nv_sata, and qlc.
The warning can be ignored.
The left pane in Packagemanager is used to display category names. These names are localized in different locales. However, the localization of the names is not changed if you switch the locale, even after restarting Packagemanager.
Click the reload button, which will reload the category names in the new locale.
This section describes desktop related issues that apply to the OpenSolaris 2009.06 OS.
After installing the OpenSolaris 2009.06 release, the hald-addon-network-discovery agent cannot be enabled. This failure occurs regardless of which method you use to enable the service.
An error message similar to the following is displayed in the service log:
[ Jul 24 21:09:32 Executing start method
("/lib/svc/method/svc-network-discovery start snmp"). ]
/usr/bin/dbus-send --system --print-reply --dest=org.freedesktop.Hal
--type=method_call /org/freedesktop/Hal/devices/network_attached
org.freedesktop.Hal.Device.NetworkDiscovery.EnablePrinterScanningViaSNMP
int32:60 string:public string:0.0.0.0
Error org.freedesktop.DBus.Error.NoReply: Did not receive a reply. Possible
causes include: the remote application did not send a reply, the message bus
security policy blocked the reply, the reply timeout expired, or the network
connection was broken.
[ Jul 24 21:09:40 Method "start" exited with status 1. ]
Use the following procedure:
Add the SUNWSmmgr package.
# pkg install SUNWSmmgr
# allow send_interface="org.freedesktop.Hal.Device.NetworkDiscovery" send_destination="org.freedesktop.Hal"/>
Restart the HAL add-on module.
- To restart the add-on module from the desktop, perform these steps:
Open the Services settings application from the main menu.
System-> Administration-> Services
The Services settings window is displayed.
- To restart the add-on module from the command line, type:
# svcadm restart hal
# svcadm disable snmp
# svcadm enable snmp
Firefox dumps core in NV build 95 due to Flash Player bug. Flash plug-in v9.0 r124 has a bug that causes Firefox to crash. This issue has been fixed in the flash plug-in v 9.0 r125.
Note - This affects only users of Flash Player v9.0 r124 and below.
Start Firefox.
From the Menu bar, click on Tools -> Add-ons -> Plugins.
Right click on Shockwave Flash and choose the Disable option to disable the plug-in.
Manually install the updated version from the following URL: http://www.adobe.com/products/flashplayer/productinfo/instructions/#section-4
When you start Tracker, the home directory is not indexed by default.
Use the following procedure:
Go to System->Preferences.
Check the Index and watch my home directory checkbox in the 'Files' tab.
Click OK.
A dialog box appears prompting that data must be reindexed.
Click the Reindex button to reindex data.
When you start Tracker, you cannot search for Evolution emails.
None.
When you maximize a window in the twinview or xinerama mode, the window fills all the screens instead of being maximized on the current screen only.
None.
Chinese, Japanese, and Korean characters do not display correctly on Java technology-based applications.
Enable the antialiasing settings manually. Use one of the following commands:
java -Dawt.useSystemAAFontSettings=on
env _JAVA_OPTIONS=-Dawt.useSystemAAFontSettings=on java
For more information about turning on the antialiasing settings, refer to Java 2D API FAQs.
This section describes hardware-related issues that apply to the OpenSolaris 2009.06 OS.
The nge driver is obtaining the ethernet address with the bytes in the reverse order from the hardware.
Uncomment the last line of the nge.conf configuration file which is located in /kernel/drv/. Reboot the machine.
This section describes localization issues that apply to the OpenSolaris 2009.06 OS.
The following locales will not be installed if your system is earlier than the OpenSolaris 2008.11 release and if the slim_install cluster is uninstalled before you upgrade your system to the OpenSolaris 2009.06 release using the pkg image-update command:
en_US.UTF-8
es_ES.UTF-8
fr_FR.UTF-8
de_DE.UTF-8
Install the missing locale packages. Execute the following commands:
If your current LANG variable is any one of the locales listed above:
$ export LANG=C
$ pfexec pkg install SUNWlang-enUS SUNWlang-frFR SUNWlang-esES SUNWlang-deDE
iiim-properties(1) help fails to load with the XML parser error in some locales.
Install SUNWgnome-xml from pkg.opensolaris.org.
$ pfexec pkg install SUNWgnome-xml
Manually execute the following script.
$ pfexec /usr/share/sgml/docbook/docbook-catalog-install.sh
The following features might not be supported in a future release of the OpenSolaris software.
This section provides troubleshooting resources for the OpenSolaris 2009.06 OS.
Terms of Use
|
Privacy
|
Trademarks
|
Copyright Policy
|
Site Guidelines
|
Site Map
|
Help
Your use of this web site or any of its content or software indicates your agreement to be bound by these Terms of Use.
© 2012, Oracle Corporation and/or its affiliates.