Sun and OracleChannel SunHow to BuyLog InPortuguês brasileiro

Solaris 10 Release Notes
  Procure somente este livro
Pesquisar Ajuda
Exibir este livro em:
Contidos dentro
Localizar Mais Documentação
Destaques de Recursos de Suporte
 Fazer download desta apostila em PDF (1069 KB)

Chapter 2 Installation Issues

This chapter describes problems that relate to the installation of the Solaris 10 Operating System.

Note –

Some of the issues and bugs in this chapter have been fixed in subsequent Solaris 10 releases. If you have upgraded your Solaris software, certain issues and bugs in this chapter might no longer apply. To see which bugs and issues no longer apply to your specific Solaris 10 software, refer to Appendix A, Table of Integrated Bugs in the Solaris 10 Operating System.

General Information

This section provides general information such as behavior changes in Solaris 10 OS.

Single UNIX Specification, Version 3 Introduces Changes

The Single UNIX Specification, Version 3 (SUSv3) provides updates to the following POSIX operating system interfaces:

  • POSIX.1-1990

  • POSIX.1b-1993

  • POSIX.1c-1996

  • POSIX.2-1992

  • POSIX.2a-1992

The updates include the following changes:

ENV Variable in Korn Shell

The environment variable ENV is no longer expanded in a noninteractive Korn shell. This change applies to both /bin/ksh and /usr/xpg4/bin/sh.

Compilers and SUSv3 Applications

Applications that conform to SUSv3 and other standards should be built by using correct feature test macros, compilers, and options. See the standards(5) man page.

file Command

When used with the -m option, the file command performs signed comparisons on user-defined magic files. Additionally, in the x86 environment, the command no longer switches multibyte integers to big-endian (SPARC) order before comparing the magic values. See the file(1) and magic(4) man pages.

Print Functions

snprintf() and vsnprintf() now return the number of bytes that would be formatted when the input value n=0. See the snprintf(3C) and vsnprintf(3C) man pages.

Text Editor

The editor checks for additional conditions before reading the .exrc files in $HOME or the current directory at startup. See the ex(1) and vi(1) man pages.

64-bit SPARC: Time Functions

If the input time_t is too large to be represented in a tm struct, the following functions return a null pointer and set errno to EOVERFLOW:

  • gmtime()

  • gmtime_r()

  • localtime()

  • localtime_r()

See the gmtime(3C), gmtime_r(3C), localtime(3C), and localtime_r(3C) man pages.

Trailing Slashes

SUSv3 allows a trailing slash character (/) in a path name only if the path specifies an existing directory or a directory to be created.

Utilities in /usr/xpg6/bin and /usr/xpg4/bin

Utilities in /usr/xpg6/bin and /usr/xpg4/bin have SUSv3-conforming or XPG4-conforming behavior that conflicts with default Solaris behavior. PATH must be properly configured for a user to use a command-line environment that conforms to SUSv3. See the man page for the specific utility for more information.

Support for TCP Wrappers is Enabled in sendmail

The utility sendmail has been added to the list of services that support TCP wrappers. The utility's added capability might cause sendmail to reject connections in Solaris 10 systems that were previously configured with very restrictive services. Use the daemon name sendmail to enable use of TCP wrappers. For information about TCP wrapper support, see sendmail Version 8.12 Uses TCP Wrappers in Solaris 10 What’s New. For information about configuring TCP wrappers, refer to the tcpd(1M) and hosts_access(4) man pages packaged in SUNWtcpd.

x86: Change to Solaris fdisk Identifier

In the Solaris 10 release, the Solaris fdisk identifier is set to 0xbf to facilitate the configuration of multiboot environments that include the Linux operating system. In previous Solaris releases, the Solaris fdisk identifier was set to 0x82, the same identifier as the Linux swap partition. The change to the identifier enables you to install the Solaris 10 OS and the Linux operating system without having to modify the Linux swap partition or reinstall the Linux OS.

The Solaris OS continues to recognize and support Solaris fdisk with the 0x82 identifier. You can use the fdisk command to switch between the 0x82 identifier and the 0xbf identifier. If you are using Solaris Live Upgrade and need to use the 0x82 identifier, the luactivate command enables you to switch the 0xbf identifier to the 0x82 identifier.

For more information, see the fdisk(1M) man page.

x86: Application Memory-Allocation Limits on x86 Systems

Solaris applications allocate memory by using functions like malloc(), brk(), or mmap(). The maximum amount of memory an application can allocate is bounded by the size of the user virtual address space. Program text, stack, data, and libraries that are contained within the user virtual address space can further reduce the maximum memory-allocation size.

32-bit Computing Environment

In a 32-bit computing environment, the 32-bit virtual address space is shared between user and kernel. Consequently, the size of the user virtual address space is 4 Gbytes minus the size of the kernel virtual address space. The configuration of the size of the kernel virtual address space depends on the size of physical memory. As the size of physical memory increases, the size of the user address space decreases. For systems with 8 Gbytes of memory, the size of the user address space is approximately 3 Gbytes.

If the memory allocation limits are insufficient for an application, two options exist:

  • Reduce the default size of the kernel virtual address space to increase the size of the user virtual address space. You can set the kernelbase variable to a higher value by using the eeprom command. Exercise caution when changing kernelbase. Otherwise, a system with a smaller kernel virtual address space might hang or might not run as efficiently.

  • As a better alternative, run the application in the Solaris x86 64-bit computing environment.

64-bit Computing Environment

In the 64-bit computing environment, the virtual address space for 32-bit applications is entirely for the user. Therefore, a 32-bit application can allocate the full 4 Gbytes minus the user address space that is needed for text, stack, and libraries.

On some early 64-bit AMD processors, the memory allocation limit can be further reduced by 1 Gbyte. To determine whether your processor is affected, perform the following steps:

  1. Run the command pmap $$ from any 32-bit shell.

  2. On the process map, check if addresses greater than or equal to 0xC0000000 are listed. If these addresses do not exist, then the Solaris OS has limited 32-bit applications to the lowest 3 Gbytes of the 4-Gbyte virtual address range.

A 64-bit application can allocate 250 Tbytes of memory depending on what the application already has in use.

x86: Solaris Supports the no execute Bit

Solaris 10 on x86 supports the no execute (NX) bit in 32-bit and 64-bit modes. The NX bit enables fully implementing the protections that are implied by not using PROT_EXEC in various memory system calls such as the mmap() function. Previously on x86, all mapped memory implicitly had PROT_EXEC set. With Solaris support for the NX bit, virus protection is enhanced.

On NX-capable hardware, Solaris OS uses the NX bit by default whenever PROT_EXEC is not specified. Stack segments, however, use PROT_EXEC by default, not the NX bit. The default setting for stack segments can be changed by performing one of the following:

  • Set no_exec_userstack in /etc/system to globally set no exec on all applications stacks.

  • Link an individual application with the /usr/lib/ld/map.noexstk map file.

The system administrator can disable all use of the NX bit by using the eeprom command to set enforce-prot-exec to “off”. This variable is provided as a transition workaround for any system with legacy applications that are missing PROT_EXEC.

Preinstallation Issues and Bugs

This section contains critical installation issues that you need to be aware of before installing or upgrading to Solaris 10 OS. These issues might have an impact that would prevent installation or upgrades from completing successfully. If bugs in this section apply to your system, you might need to perform the recommended workarounds before you install or upgrade.

Support for Products Not Part of the Solaris OS

Although the Solaris 10 software has been tested for compatibility with previous releases, some third-party applications might not be fully ABI compliant. Contact the supplier of these applications directly for information about compatibility.

Your system might run both a Solaris OS and other products that are not part of the Solaris software. These products might be supplied by either Sun or another company. If you upgrade this system to the Solaris 10 release, make sure that these other products are also supported on the Solaris 10 OS. Depending on the status of each of these products, you can perform one of the following options:

  • Verify that the existing version of the product is supported on the Solaris 10 software.

  • Install a new version of the product that is supported on the Solaris 10 release. You might need to remove the previous version of the product prior to upgrading to the Solaris software. See the product documentation for more details.

  • Remove the product prior to upgrading to the Solaris 10 software.

Information About Solaris 10 3/05 HWx Releases

The Solaris 10 3/05 HWx releases are temporary, limited releases and are being replaced by the Solaris 10 1/06 software. The information in this section is useful only to those customers whose platforms initially required HW1 or HW2, but who have not yet updated to the Solaris 10 1/06 OS.

Solaris 10 3/05 HW1 OS Must Be Installed on Certain Systems

The Solaris 10 3/05 HW1 Operating System is based on the Solaris 10 3/05 OS, but has support added for specific new Sun hardware. Install the Solaris 10 3/05 HW1 OS only on systems or domains that use hardware products that require the OS. For all other systems or domains, use the original Solaris 10 3/05 software.

To determine whether your hardware requires Solaris 10 3/05 HW1 OS, see the documentation for that hardware. If your system includes UltraSPARC IV+ boards, you might need additional information to use the Solaris 10 3/05 HW1 OS on that hardware. The additional information is in the Solaris 10 Sun Hardware Platform Guide at http://www.sun.com/products-n-solutions/hardware/docs/Software/Solaris_on_SunHardware/Solaris_10/index.html.

The following information applies only to platforms still running the temporary Solaris 10 3/05 HW1 release. The information does not apply to systems that are running the Solaris 10 3/05, Solaris 10 3/05 HW2, or Solaris 10 1/06 OS.

Solaris Install Launcher Exiting Causes Reboots (6300863)

In some situations, installation of Solaris 10 3/05 HW1 fails when the Solaris Install Launcher fails to find find_device.out, reboots, then reaches the same point and prompts for a reboot again.

Workaround: Specify the -text option when booting:


boot net -text

Race Condition Between EF/kcfd and IPsec Algorithm Availability (6266083)

This problem might occur on a freshly installed system or a system that imports a large number of new Service Management Facility (SMF) manifests during the boot. After these booting conditions, IPsec, which is part of svc:/network/initial:default, might be initialized prior to the encryption framework, which is part of svc:/system/cryptosvc:default. Because authentication or encryption algorithms are not available, creation of IPsec security associations might fail with an error message such as the following:


PF_KEY error: type=ADD, errno=22:
Invalid argument, diagnostic  code=40:
Unsupported authentication algorithm

For example, this error might occur when using DR on a Sun Fire E25K system, which involves IPsec services.

Workaround: Before performing operations that use IPsec services, perform the following steps after a boot that imports a large number of new SMF manifests:

  1. Issue this command after booting:


    ipsecalgs -s

  2. If /etc/inet/secret/ipseckeys exists on the system, also issue this command:


    ipseckey -f /etc/inet/secret/ipseckeys

Now you can perform actions that create IPsec security associations, such as using DR on a Sun Fire E25K system.

This procedure needs to be repeated only when a large number of new SMF manifests are imported during the boot.

Solaris 10 3/05 HW2 OS Must Be Installed on Certain Systems

The Solaris 10 3/05 HW2 Operating System is the Solaris 10 3/05 OS with additional patches to support the Sun FireTM T2000 System, which features the new Sun UltraSPARC T1 processor. The HW2 OS is required on the T2000 system, and should not be installed on any other systems. For more information see the Sun Fire T2000 Product Notes and the Solaris 10 Sun Hardware Platform Guide.

Note –

The HW2 OS is not a replacement for either the Solaris 10 3/05 or Solaris 10 3/05 HW1 operating systems. All Sun systems except those based on the Sun UltraSPARC IV+ and T1 processors should be running the Solaris 10 3/05 OS.

Upgrading From Solaris Express or Solaris 10 Beta Releases Not Supported

Note –

For the latest information about upgrade support beginning with the Solaris 10 1/06 release, see Changes in Upgrade Support for Solaris Releases.

Upgrading to Solaris 10 OS is supported only if you are upgrading from one of these releases:

  • Solaris 9 9/04

  • Solaris 8 2/04 (SPARC Platform Edition)

  • Solaris 8 2/02 (Intel Platform Edition)

  • Solaris 7 11/99

Upgrading from any Solaris Express or Solaris 10 Beta release, whether by using standard upgrade or Live Upgrade, is not supported. To upgrade from these releases, you must install Solaris 10 OS.

Some Executable Files Not Signed After You Upgrade to Solaris 10 OS

Most executable files in Solaris 10 contain digital signatures that can be used to verify the files' authenticity. See Signing ELF Objects in Solaris 10 What’s New for information about this feature.

During an upgrade from a Solaris Express or Solaris 10 Beta release, some files might not be replaced with digitally signed versions. The missing signatures have no effect on the ability to run the programs, but the authenticity of the files cannot be verified. To prevent this problem, perform a full installation of Solaris 10 OS instead of an upgrade. See Upgrading From Solaris Express or Solaris 10 Beta Releases Not Supported.

Sun Fire V250 Server Installation

The SUNWCXall software metacluster must be installed on a Sun Fire V250 Server.

NFS Version 4 Introduces New Prompt at First System Boot

Solaris 10 OS uses NFS Version 4 (NFSv4) as the default distributed file system mechanism. In NFSv4, file user and group attributes are exchanged as strings of the form “user@domain” and “group@domain,” respectively.

The domain portion of the string should be common between client and server. This information is automatically derived from the system's name service domain name. However, in certain cases, the derived domain might not match between client and server. This mismatch might typically occur in environments where either of the following circumstances is true:

  • The DNS TXT record is not used.

  • Multiple administrative domain boundaries are crossed.

If domain information between client and server does not match, NFSv4-accessed files might appear to be owned by “nobody.”

To ensure proper configuration, on the first system boot the system now prompts you for a domain to use for NFSv4. Answer “no” to this prompt if one of the following configurations describes your deployment:

  • Your deployment utilizes the DNS TXT record to configure the NFSv4 domain.

  • Your deployment has a single (flat) administrative domain.

Custom JumpStart and Advanced Installations

Environments that use the Custom JumpStartTM method to perform unattended system installations require minimal tuning to suppress the first boot prompts.

A sample script, set_nfs4_domain, is provided in the Solaris 10 distribution media and is located in the ${CDrom_mnt_pt}/Solaris_10/Misc/jumpstart_sample directory. A copy of this script should be modified to set the NFS4_DOMAIN variable to site-specific needs and be called from within JumpStart's finish script. The target system's NFSv4 domain is therefore preconfigured as part of the JumpStart finish phase, and thus suppresses any first boot prompts.

Follow these steps to perform unattended installations:

  1. Create a copy of the set_nfs4_domain script in the same directory as your JumpStart's finish.sh scripts.

  2. Edit the script and set the NFS4_DOMAIN variable to your specific needs.

  3. Edit your finish.sh script and append a call to set_nfs4_domain.

  4. Perform the JumpStart installations as usual.

For further details, refer to the sysidconfig(1M), sysidtool(1M), nfsmapid(1M), and nfs(4) man pages. See also the System Administration Guide: Network Services.

x86: Systems With elx or pcelx NICs Fail Network Configuration

Systems with an elx or a pcelx network interface card (NIC) fail to install. During the configuration of the NIC, the following error message might be displayed:


WARNING: elx: transmit or jabber underrun: d0<UNDER, INTR, CPLT>

See the elxl(7D) or pcelx(7D) man page for more information.

Workaround: Install and run on systems that do not have elx or pcelx NICs.

Default Size of /var File System Inadequate for Extra Value Products

The default size of the /var file system might be insufficient for the Extra Value products in the following situations:

  • If you install any of the Extra Value products that are provided on the Solaris 10 DVD or CDs

  • If the /var file system is located on a separate slice

You must manually specify a larger slice size for the /var file system.

Note –

If the /var file system is not on a separate slice or partition, this problem does not occur.

Workaround: Choose one of the following workarounds.

  • If you are using the Solaris installation program GUI, follow these steps.

    1. Begin the installation.

    2. From Select Type of Install, select Custom Install.

      The Solaris installation program displays several screens that enable you to customize the software localizations, products, and disk layout that you want to install.

    3. From Lay Out File Systems, select Modify.

      The disk layout screen is displayed.

    4. Type /var in the File System column for a specific slice, then click Apply.

      The installation program suggests a default size for the /var file system.

    5. Edit the Size column entry for the /var file system to twice the disk space size.

      For example, if the installation program assigns 40 Mbytes of space, change the Size value to 80.

    6. Complete the installation.

  • If you are using the Solaris installation program's text installer, follow these steps.

    1. Begin the installation.

    2. From Select Type of Install, select Custom Install.

      The Solaris installation program displays several screens that enable you to customize the software localizations, products, and disk layout that you want to install.

    3. From Lay Out File Systems, select Auto Layout.

      The disk layout screen is displayed.

    4. Type /var in the File System column for a specific slice.

      The installation program suggests a default size for the /var file system.

    5. Press F4_Customize to customize the size of the /var file system.

    6. Edit the Size column entry for the /var file system to twice the disk space size.

      For example, if the installation program assigns 40 Mbytes of space, change the Size value to 80.

    7. Complete the installation.

  • If you are using the custom JumpStart program, use the filesys profile keyword to set the size of the /var file system. The following example sets the size of the /var file system on slice 5 to 256 Mbytes.


    filesys  c0t0d0s5 256 /var

x86: Do Not Upgrade Hewlett-Packard (HP) Vectra XU Series Systems With BIOS Version GG.06.13

The Solaris 10 software includes a feature that enables you to install large partitions. The system BIOS must support logical block addressing (LBA). BIOS Version GG.06.13 does not support LBA access. The Solaris boot programs cannot manage this conflict. This issue can also affect other HP Vectra systems.

If you perform this upgrade, your HP system can no longer boot. Only a blank black screen with a flashing underscore cursor is displayed.

Workaround: Do not upgrade HP Vectra XU Series systems with the latest BIOS Version GG.06.13 to the Solaris 10 release. This version no longer supports these systems.

You can still boot your system by using the boot diskette or boot CD because the boot paths do not use the hard disk code. Then select the hard disk as your bootable device instead of the network or CD-ROM drive.

SPARC: Older Firmware Might Need Boot Flash PROM Update

On SPARC based systems, Solaris 10 OS runs in 64–bit mode only. Some Sun4UTM systems might need to be updated to a higher level of OpenBootTM firmware in the flash PROM to run the OS in 64-bit mode. The following systems might require a flash PROM update:

  • UltraTM 2

  • Ultra 450 and Sun EnterpriseTM 450

  • Sun Enterprise 3000, 4000, 5000, and 6000 systems

The following table lists the UltraSPARC systems and the minimum firmware versions that are required to run the 64–bit Solaris 10 OS. System type is the equivalent of the output of the uname -i command. You can determine which firmware version you are running by using the prtconf -V command.

Table 2–1 Minimum Firmware Versions Required to Run 64–Bit Solaris Software on UltraSPARC Systems

System Type From uname -i

Minimum Firmware Version From prtconf -V

SUNW,Ultra-2 

3.11.2 

SUNW,Ultra-4 

3.7.107 

SUNW,Ultra-Enterprise 

3.2.16 

Note –

If a system is not listed in the previous table, the system does not need a flash PROM update.

See any edition of the Solaris 8 Sun Hardware Platform Guide at http://docs.sun.com for instructions to perform a flash PROM update.

Additional Patches Are Needed to Run Solaris Live Upgrade

For Solaris Live Upgrade to operate correctly, a limited set of patch revisions must be installed for a given OS version. Make sure you have the most recently updated patch list by consulting http://sunsolve.sun.com. For additional information, search for the info Doc 72099 on the SunSolveSM web site.

Limitation When Installing Solaris Live Upgrade Packages

If you are running the Solaris 7 or Solaris 8 release, you might not be able to run the Solaris Live Upgrade installer. These releases do not contain the set of patches that is needed to run the Java 2 runtime environment.

The typical failure that results is a Java exception error. The following messages might be displayed:


InvocationTargetException in ArchiveReader constructornull
          java.lang.reflect.InvocationTargetException
                 at install.instantiateArchiveReader(Compiled Code)
                 at install.<init>(Compiled Code)
                 at install.main(Compiled Code)

To run the Solaris Live Upgrade installer and install the packages, you must have the Java 2 runtime environment recommended patch cluster.

Workaround: Complete the following workaround:

Solaris Management Console 2.1 Software Is Not Compatible With Solaris Management Console 1.0, 1.0.1, or 1.0.2 Software

Solaris Management Console 2.1 software is not compatible with Solaris Management Console 1.0, 1.0.1, or 1.0.2 software. If you are upgrading to the Solaris 10 release, and you have Solaris Management Console 1.0, 1.0.1, or 1.0.2 software installed, you must first uninstall the Solaris Management Console software before you upgrade. Solaris Management Console software might exist on your system if you installed the SEAS 2.0 overbox, the SEAS 3.0 overbox, or the Solaris 8 Admin Pack.

Workaround: Choose one of the following workarounds:

  • Before you upgrade, use the /usr/bin/prodreg command to perform a full uninstall of Solaris Management Console software.

  • If you did not uninstall Solaris Management Console 1.0, 1.0.1, or 1.0.2 software before you upgraded to the Solaris 10 release, you must first remove all Solaris Management Console 1.0, 1.0.1, or 1.0.2 packages. Use the pkgrm command for package removal instead of the prodreg command. Carefully follow the order of package removal. Complete the following steps:

    1. Become superuser.

    2. Type the following command:


      # pkginfo | grep “Solaris Management Console“

      If the description does not start with “Solaris Management Console 2.1,” the package names in the output identify a Solaris Management Console 1.0 package.

    3. Use the pkgrm command to remove all instances of Solaris Management Console 1.0 packages in the following order:

      Note –

      Do not remove any package that has “Solaris Management Console 2.1” in its description. For example, SUNWmc.2 might indicate Solaris Management Console 2.1 software.

      If the pkginfo output displays multiple versions of Solaris Management Console 1.0 packages, use the pkgrm command to remove both packages. Remove the original package. Then, remove the package that has been appended with a number. For example, if the SUNWmcman and SUNWmcman.2 packages appear in the pkginfo output, first remove the SUNWmcman package and then remove the SUNWmcman.2 package. Do not use the prodreg command.


      # pkgrm SUNWmcman
# pkgrm SUNWmcapp
# pkgrm SUNWmcsvr
# pkgrm SUNWmcsvu
# pkgrm SUNWmc
# pkgrm SUNWmcc
# pkgrm SUNWmcsws

    4. In a terminal window, type the following command:


      # rm -rf /var/sadm/pkg/SUNWmcapp

    The Solaris Management Console 2.1 software should now function properly. For future maintenance, or if the Solaris Management Console 2.1 software does not function properly, remove the Solaris Management Console 2.1 software. Reinstall the software by completing the following steps:

    1. Use the pkgrm command to remove all Solaris Management Console 2.1 packages and dependent packages in the following order:

      Note –

      If your installation has multiple instances of Solaris Management Console 2.1 packages, such as SUNWmc and SUNWmc.2, first remove SUNWmc, and then SUNWmc.2 . Do not use the prodreg command.


      # pkgrm SUNWpmgr 
# pkgrm SUNWrmui 
# pkgrm SUNWlvmg 
# pkgrm SUNWlvma 
# pkgrm SUNWlvmr 
# pkgrm SUNWdclnt 
# pkgrm SUNWmga 
# pkgrm SUNWmgapp 
# pkgrm SUNWmcdev 
# pkgrm SUNWmcex 
# pkgrm SUNWwbmc 
# pkgrm SUNWmc 
# pkgrm SUNWmcc 
# pkgrm SUNWmccom

    2. Insert the Solaris 10 Software - 4 CD into your CD-ROM drive. Type the following in a terminal window:


      # 
# cd /cdrom/cdrom0/Solaris_10/Product
# pkgadd -d . SUNWmccom SUNWmcc SUNWmc SUNWwbmc SUNWmcex SUNWmcdev \
  SUNWmgapp SUNWmga SUNWdclnt SUNWlvmr SUNWlvma SUNWlvmg SUNWpmgr \
  SUNWrmui

    All previous Solaris Management Console versions are removed. The Solaris Management Console 2.1 software is now functional.

Installation Fails When You Install Solaris Flash Archive on Empty Boot Environment With Solaris Live Upgrade (6222925)

If you use Solaris Live Upgrade to install a Solaris Flash archive on an empty alternate boot environment, the installation fails. Messages that are similar to the following errors are displayed.


sh:  /path_to_install_image/Solaris_10/Tools/Boot/usr/sbin/install.d
/mergescripts/merge_name_service:  not found

ERROR: Transfer list entry (/a/var/svc/profile/name_service.xml):
mergescript failed (/path_to_install_image/Solaris_10/Tools/
Boot/usr/sbin/install.d/mergescripts/merge_name_service)

ERROR: Unable to copy a temporary file to it's final location

ERROR: Flash installation failed

Extracting Flash Archive: 100% completed (of 4361.24 megabytes)
The Solaris flash install of the BE S10B74L2 failed.

In the previous example, path_to_install_image is the directory path to the network installation image you are using for the installation.

Workaround: Follow these steps.

  1. Change directories to the mergescripts directory in the network installation image that you are using for the installation.


    # cd path_to_install_image/Solaris_10/Tools
/Boot/usr/sbin/install.d/mergescripts

  2. Create the merge_name_service file.


    # touch merge_name_service

  3. Make the merge_name_service file executable.


    # chmod a+x merge_name_service

  4. On the system you want to install, install the Solaris Flash archive on the alternate boot environment with Solaris Live Upgrade.

    For instructions about how to install a Solaris Flash archive with Solaris Live Upgrade, see Installing Solaris Flash Archives on a Boot Environment in Solaris 10 Installation Guide: Solaris Live Upgrade and Upgrade Planning.

  5. Mount the file systems in the alternate boot environment you installed in step 4.


    # lumount alt_boot_env 
mount-point

    In the previous command, alt_boot_env is the boot environment you installed in step 4. mount-point is the mount point for the alternate boot environment.

  6. Change directory to the /var/svc/profile directory in the alternate boot environment.


    # cd mount-point/var/svc/profile

  7. Remove the name_service.xml profile.


    # rm name_service.xml

  8. Create a symbolic link for the name service profile to the name_service.xml file.


    # ln -s ns_name-service.xml name_service.xml

    In the previous example, name-service is the name service you want to use for the system. Use one of the following strings for this value.

    • nis

    • nisplus

    • dns

    • ldap

    • files

Solaris GUI Installation Program Fails If You Configure Nonprimary Interface and Enable DHCP (6215739)

If you try to configure a nonprimary board interface during a Solaris 10 installation, the installation fails under the following conditions.

  • You use the Solaris GUI installation program.

  • You try to configure a nonprimary interface, such as a PCMCIA interface card.

  • You enable the system to use Dynamic Host Configuration Protocol (DHCP).

The following error message is displayed.


An error has occurred in configuring the system.
Please go back and check your settings.

Workaround 1: Use the Solaris Interactive Text (Desktop Session) installer.

  • For SPARC based systems, type the following command at the ok prompt.


    ok boot cdrom -text

  • For x86 based systems, type the following command when the system prompts you to select an installation type.


    b -text

Workaround 2: Use the Solaris installation GUI program.

  1. When the installation program prompts your for Network Connectivity information, select Non-networked.

  2. Complete the installation.

  3. After the installation, use the sys-unconfig command to reconfigure the system as a networked system with DHCP support enabled.

    For more information about reconfiguring a system with the sys-unconfig command, see the sys-unconfig(1M) man page.

x86: Sun Java Workstations 2100Z Might Panic When Booting From Solaris 10 Operating System DVD (6214356)

The DVD combo-drive firmware in a Sun Java Workstation 2100Z might cause a system panic. The panic occurs when you boot the workstation from the Solaris 10 Operating System DVD. After the kernel banner is displayed, the following message is very quickly flashed:


panic[cpu0]/thread=fec1be20: mod_hold_stub: 
Couldn't load stub module sched/TS_DTBL
fec25cb0 genunix:mod_hold_stub+139 (fec04088, 63, fea11)
fec25cc4 unix:stubs_common_code+9 (1, 8, fec026e4)
fec25ce4 unix:disp_add+3d (fec026dc)
fec25d00 genunix:mod_installsched+a4 (fef01530, fef01518)
fec25d20 genunix:mod_install+2f (fef01518, fec25d3c,)
fec25d2c TS:_init+d (0, d6d89c88, fec25d)
fec25d3c genunix:modinstall+d9 (d6d89c88)
fec25d50 genunix:mod_hold_installed_mod+2e (d6d77640, 1, fec25d)
fec25d7c genunix:modload+ac (fec026c4, fec26c4)
fec25d98 genunix:scheduler_load+3d (fec026c4, fec026dc)
fec25db4 genunix:getcid+50 (fec026c4, fec28514)
fec25dcc unix:dispinit+df (fec25ddc, fe814ba9)
fec25dd4 unix:startup_modules+d5 (fec25dec, fe8cac37)
fec25ddc unix:startup+19 (fe800000, 166130, 7)
fec25dec genunix:main+16 ()

Then the system automatically resets.

Workaround: Choose one of the following options:

Workaround 1: Modify some BIOS configuration settings. This temporary workaround enables a Solaris 10 installation to be completed. However, this method might cause poor read-DVD performance. Follow these steps:

  1. During system boot, press F2 at the prompt to enter the setup.

    The screen displays attachment-type options similar to the following example:


    Primary Master [ ]
Primary Slave [ ]
Secondary Master [CD-ROM]
Secondary Slave [ ]

  2. Choose the DVD drive's attachment type by selecting the attachment type for CD-ROM.

    Note –

    The screen might display more than one attachment type for CD-ROM, for example, if your system has multiple optical drives. In such cases, you might need to open the system case to determine the DVD drive's point of attachment. Make sure that you select the correct attachment type that applies to the DVD drive.

  3. After selecting the correct CD-ROM attachment type, press Enter.

    The next screen appears with Type:[Auto] automatically selected.

  4. Press the spacebar twice to change the selection to Type:[CD-ROM].

  5. Use the arrow keys to select Transfer Mode.

  6. Press Enter to display a list of other Transfer Mode options.

  7. Use the arrow keys to select Standard, then press Enter to accept your selection.

  8. Press F10 to save the configuration changes and exit BIOS setup.

    The system restarts.

Workaround 2: Update the DVD combo drive's firmware to v1.12. This option requires your DVD combo drive to be attached to a system that is running Microsoft Windows. Follow these steps.

  1. Remove your DVD combo drive from the Sun Java Workstation 2100z. See the workstation's user's guide for steps to remove the drive properly.

  2. Connect the drive to a system that is running Microsoft Windows. Make sure to change the drive's master and slave jumper settings, if needed.

  3. Go to AOpen's download center at http://download.aopen.com.tw/default.aspx.

  4. Search for your DVD drive's firmware by using the following information:

    • Product: Combo drives

    • Model: COM5232/AAH

    • Categories: Firmware

  5. Download and install the firmware version R1.12.

  6. Reinstall the drive on the workstation. If needed, restore the original master and slave jumper settings.

Note –

Newer versions of the firmware might already be available at the site. Sun's tests confirm that the v1.12 release resolves the panic issue. Sun cannot confirm whether newer firmware revisions after v1.12 similarly resolve the problem.

x86: Solaris Installation GUI Might Fail When You Install Solaris Flash Archive (6208656)

If you use the Solaris installation graphical user interface (GUI) to install a Solaris FlashTM archive on an x86 based system, the installation might fail. This failure occurs on systems that are not running the Solaris OS or the Linux operating system before you begin the installation.

The following error message is displayed.


Unexpected error found with media or archive selected

Workaround: When the Solaris installation program prompts you to select an installation type, do not select the Solaris Interactive option. Select either option 3 or 4 and perform a Solaris Interactive Text installation. When prompted to specify the media to use during the installation, type the path to the Solaris Flash archive.

x86: Serial Consoles of Some Sun Fire Systems Do Not Work (6208412)

The serial console (ttya) on the following Sun Fire systems does not work by default:

  • Sun Fire V20z

  • Sun Fire V40z

  • Sun Fire V60x

  • Sun Fire V65x

To use the serial console, you must manually configure the system's BIOS.

Workaround: This workaround requires your system to have a Sun keyboard and a display monitor. Follow these steps:

  1. Boot the machine.

  2. During system boot, press F2 at the prompt to access the Phoenix BIOS.

  3. Under Peripherals, change the comm port from disabled to enabled.

  4. Save the configuration and boot the system.

  5. Use the eeprom command to change input-device and output-device to ttya.

Note –

Pressing the Stop and N keys at system boot to reset the low-level firmware to default settings does not work on these systems.

Solaris Installation GUI Program Might Fail on Systems With Existing x86 fdisk Boot Partitions (6186606)

The Solaris installation GUI program might fail on a system with an existing x86 boot partition. The failure occurs if the existing x86 boot partition was created with the Solaris text-based installer. The following error message is displayed.


Default layout will not work on this system.
Error:
Error: ERROR: Could not create explicit fdisk partition on c0t0d0,
requested cylinders 14581 - 14597 in use by fdisk partition 1
Error:
Error: ERROR: System installation failed
Pfinstall failed. Exit stat= java.lang.UNIXProcess@a89ce3 2 
artition on c0t0d0, requested cylinders 14581 - 14597 in use by fdisk
partition 1  ERROR: System installation failed

Workaround: Choose one of the following workarounds.

Workaround 1: When the installation program prompts you to select an installation type, select 3 Solaris Interactive Text (Desktop Session).

Workaround 2: If you use the Solaris installation GUI program, follow these steps.

  1. Begin the installation.

  2. At the prompt to select an installation type, select Custom Install.

    The custom installation panels prompt you for information about the locales, software, and disks that you want to install.

  3. Answer the questions on the screens as appropriate for your system.

  4. On the Fdisk Selection screen, check the disk that contains the x86boot partition.

  5. Remove the x86boot partition by changing it to UNUSED in the pull-down menu.

  6. Add the x86boot partition back by changing UNUSED back to x86boot.

  7. Continue the installation.

x86: Only Part of the Disk Is Usable by fdisk or format Commands (5042195)

On x86 based systems, when a disk device exceeds 65,535 cylinders, only the lower 16-bits of the cylinder count are used by the Solaris OS. Consequently, the actual size of disk space available for the Solaris OS is reduced. For example, the reported number of cylinders for a disk is 70,000. The difference between 70,000 and 65,535 equals the number of available cylinders for the Solaris OS, which is 4,464.

To verify whether your disk has exceeded 65,535 cylinders, compute your disk's approximate capacity by using the following formula:

disk capacity = cylinders * heads * sectors * 512

Then, compare the results with the stated capacity of your disk. A large difference between the two values indicates that your disk has exceeded 65,535 cylinders.

Workaround: Choose from the following options:

  • Keep RAID volumes small enough to avoid exceeding the cylinder limit.

  • Create a Solaris label where the geometry uses cylinders fewer than 65,535. You implement this second option by increasing the value for sectors or heads, neither of which should exceed 16 bits.

    The second option is applicable on a Sun Fire V20z or Sun Fire V40z system that has a 72-Gbyte disk mirrored. The BIOS RAID function would report a much smaller number of heads and sectors than the single 72-Gbyte disk. Consequently, the number of cylinders exceeds the 65,535 limit.

    However, you can format and label the first disk with Solaris 10 OS, and then have the BIOS mirror the first disk to another disk. Through this step, the label of the first disk is preserved. Solaris uses the geometry of the first disk and avoids exceeding the limit.

    You can also achieve the second option by using a disk editor to manually modify the label on a disk by using valid geometry values.

x86: X Server Unable to Open Mouse Device in Sun LX50 Servers (5027771)

On a Sun LX50 system, the X server might be unable to open a PS/2 mouse device. The error might occur either during installation or when the system reboots. When the problem occurs, the following error message is displayed:


ddxSUNWmouse: Error opening mouse device '/dev/kdmouse;
     /dev/kdmouse: No such device or address

Consequently, the Solaris installation program can proceed only in command-line interface mode. After installation, the problem might persist during Solaris runtime.

Workaround: On the Y-cable that connects to the server's PS/2 connector, switch the connections between the PS/2 keyboard and mouse.

If the PS/2 mouse remains unrecognized after a system reboot, press the LX50 server's Reset button. Alternatively, use the reconfigure option (b -r) when the following boot prompt appears:


Select (b)oot or (i)nterpreter:

Cannot Access Data on Solaris 10 Operating System DVD When Running Solaris 7 Software (4511090)

If your system is running the Solaris 7 software, Volume Management incorrectly mounts the Solaris 10 Operating System DVD. The DVD can be mounted, but the data is inaccessible. As a result, you cannot set up an install server, perform a Live Upgrade, or access any data on the media.

Workaround: Choose one of the following workarounds:

  • Apply the following patches:

    • 107259–03

    • 107260–03

  • Mount the Solaris 10 Operating System DVD manually. Do not use Volume Management to mount the DVD. Follow these steps:

    1. Become superuser.

    2. Stop Volume Management.


      # /etc/init.d/volmgt stop

    3. Mount the Solaris 10 Operating System DVD manually.


      # mkdir /mnt1
# mount -F hsfs -o ro /dev/dsk/c0t6d0s0 /mnt1

    4. Verify that the DVD is mounted. Then verify that the data is accessible.


      # cd /mnt1 # ls

      The system returns the following information if the DVD is correctly mounted.


      Copyright Solaris_10

Installation Bugs

The following bugs might occur during or after the installation of Solaris 10 OS.

Java Error Messages Are Displayed After a Solaris 10 OS Installation (6218158)

When the system reboots after a Solaris 10 OS installation is completed, Java error messages similar to the following example might be generated. These messages might either be displayed on the console or recorded in the /var/adm/messages file.


java[16552]: [ID 874347 user.error] libpkcs11:
open /var/run/kcfd_door: No such file or directory
java[16552]: [ID 523743 user.error] libpkcs11:
/usr/lib/security/pkcs11_softtoken.so
unexpected failure in ELF signature verification.
System may have been tampered with.
Cannot continue parsing /etc/crypto/pkcs11.conf
java[16580]: [ID 874347 user.error] libpkcs11:
open /var/run/kcfd_door: No such file or directory
java[16580]: [ID 523743 user.error] libpkcs11:
/usr/lib/security/pkcs11_softtoken.so
unexpected failure in ELF signature verification.
System may have been tampered with.
Cannot continue parsing /etc/crypto/pkcs11.conf

Workaround: None. Ignore the error messages.

Solaris 10 Installation Disc Ejects When You Install the Solaris Flash Archive (6215847)

During installation, the system might eject the disc and the installation might fail. This error might occur if you boot from a Solaris 10 installation disc. The installation disc can either be the Solaris 10 Operating System DVD or the Solaris 10 Software - 1 CD. After you boot the system, you attempt to install a Solaris Flash archive from the network.

The problem occurs under the following two sets of circumstances.

  • While running the Solaris GUI installation program from the DVD or CD installation disc, you specify an invalid Solaris Flash archive for your system. For example, specifying a path to an x86 archive while installing on a SPARC based system would cause the error. After the disc ejects, messages similar to the following errors might be displayed in the console window.


    /sbin/install-solaris: rm: cannot execute
/sbin/install-solaris: date: cannot execute
/sbin/install-solaris: gettext: cannot execute.

  • You run the Solaris text installer in a console session from the Solaris 10 installation disc and successfully install a Solaris Flash archive from the network. After the disc ejects, messages similar to the following errors might be displayed in the console window.


    /sbin/setup-launcher[310]: 1356 Killed
/sbin/setup-launcher[315]: 1357 Killed

Workaround 1: If the error was caused by an invalid archive, complete one of the following steps, depending on the error triggers.

  • Restart the installation with the GUI installation program. Then specify a path to a valid archive.

  • Install the archive by using the text-based installer in a desktop windowing session:

    • For SPARC based systems, type boot cdrom -text at the ok prompt.

    • For x86 based systems, at the prompt to choose an installation type, select 3 Solaris Interactive Text (Desktop session).

Workaround 2: If the error occurred after you used the text-based installer in a console session, choose one of the following options:

  • Use the installer in a desktop windowing session instead:

    • For SPARC based systems, type boot cdrom -text at the ok prompt.

    • For x86 based systems, at the prompt to choose an installation type, select 3 Solaris Interactive Text (Desktop session).

  • Perform the following steps after the system ejects the installation disc.

    1. Reinsert the DVD or CD in the machine.

      • If the system responds to the disc insertion, allow the system to complete the installation.

      • If the system does not respond to the installation disc, go to Step 2.

    2. Manually reboot the system.

    3. If prompted, reboot the system.

      The installation is complete.

x86: kdmconfig Program Runs Twice After Initial Installation (6209092)

During the first system reboot after an initial Solaris 10 OS installation, the kdmconfig program might run twice. During this initial startup, each time the program starts, you are prompted to select an X server. This situation occurs on x86 based systems that have not been configured to use the X server. An example of such a system is a “headless” system that is not using a graphics display.

Workaround: Each time the program starts and prompts for an X server, press F3_Exit. A short delay might occur before the startup process resumes. After this initial startup is completed, the problem no longer occurs if the system is subsequently rebooted.

x86: System Fails to Boot After Custom JumpStart Installation (6205478)

If you use the Custom JumpStart installation method to install on an x86 based system, and explicitly configure slice 2 as the overlap slice in the profile, an error occurs. The system does not reboot successfully after the installation is completed. The following error message is displayed:


Cannot find Solaris partition

This failure occurs because the overlap slice 2 (c0t0d0s2, for example) is set to begin at cylinder 1 rather than cylinder 0.

Workaround: In the Custom JumpStart profile, remove the filesys keyword entry that configures slice 2 as the overlap slice. For example, you would remove a keyword entry that is similar to the following entry.


filesys c0t0d0s2 all overlap

After you remove the entry, perform the Custom JumpStart installation.

Installation Logs Might Be Incomplete or Inaccurate (5087588)

If you install the Solaris 10 software, the install_log file in /var/sadm/system/logs and /var/sadm/install/logs might be incomplete. These logs might not include information about packages that were successfully added to the system, and they might not record errors encountered when adding packages. Consequently, you cannot use the install_log file as a resource to troubleshoot installation or application issues.

Workaround: If an application does not run correctly, do not check the install_log file to determine if the application packages were installed. Follow these steps.

  1. Check the Solaris Product Registry to determine which packages the application uses.


    # prodreg

  2. Use the pkginfo command to determine if any of these packages are partially installed.


    # pkginfo -p package-name

  3. For each package that is reported in the output of the previous command, remove the package.


    # pkgrm options package-name

  4. Reinstall each package.


    # pkgadd options package-name

x86: GUI Interactive Installation From DVD Fails if boot-device Variable Is Not Set (5065465)

If you install Solaris from the Solaris 10 Operating System DVD, the interactive GUI installation might fail. This failure occurs if the boot-device configuration variable is not set on the system.

To determine if the boot-device configuration variable is set, type the following command.


# prtconf -pv | grep boot-device

If the output of this command is boot-device: with no associated device, you cannot use the interactive GUI installation program to install from the Solaris 10 Operating System DVD.

Workaround: Use the interactive text installer to install Solaris 10 software. When the installation program asks you to select a type of installation, select option 3, Solaris Interactive Text (Desktop Session).

For more information about installing from the Solaris 10 Operating System DVD, see Solaris 10 Installation Guide: Basic Installations.

SPARC: Solaris 10 OS Installation Program Might Not Display Special Case Panels Properly (5002175)

If you install the Solaris 10 software by using the Solaris installation program, some of the information and exit panels for the SunSM Remote Services (SRS) Net Connect software might not display properly.

The installation panel frame and Cancel button are displayed, but the installation panel content is missing.

Note –

This problem might also occur during a Solaris installation of other software products that are included in this Solaris OS release.

Workaround: Follow these steps:

  1. Bypass the SRS Net Connect installation by clicking the Cancel button when the empty installation panel is displayed.

  2. After the installation of the Solaris 10 software has completed, manually install the SRS Net Connect software from the Solaris 10 Operating System DVD or Solaris 10 Software - 2 CD.

Installing the SRS Net Connect software after the initial Solaris 10 software installation has completed ensures that all panels are displayed properly.

x86: USB Keyboards Might Freeze During Install on Some Dell Precision Workstations (4888849)

During installation, USB keyboards on some Dell Precision Workstations might freeze or become partially inoperative, thereby preventing installation.

Workaround: Perform one of the following workarounds:

  • Switch the USB emulation mode in the firmware.

  • Switch to a PS/2 keyboard.

  • Reboot and try again.

Alternatively, update the system's BIOS to resolve the problem.

/dev and /devices/pseudo Permissions Set Incorrectly After Installation (4720192, 6215918)

When you install the Solaris 10 OS, the /devices/pseudo namespace permissions might be set incorrectly. The system might not be able to access some system devices. Similar namespace-permission errors occur if you installed the Solaris 10 OS on a clone system with a Solaris Flash archive.

If you run the pkgchk -n command on the SUNWcsd package after the installation, a message similar to the following output is displayed.


# pkgchk -n SUNWcsd
ERROR: /devices/pseudo/cvc@0:cvc
 pathname does not exist
 unable to create character-special device
ERROR: /devices/pseudo/cvcredir@0:cvcredir
 pathname does not exist
 unable to create character-special device

If you installed the Solaris 10 OS on a clone system with a Solaris Flash archive, the previous command produces errors similar to the following output.


# pkgchk -n SUNWcsd
ERROR: /dev
      group name sys expected <(null)> actual
ERROR: /devices/pseudo/clone@0:eri
      permissions <0666> expected <0600> actual
ERROR: /devices/pseudo/clone@0:ibd
      permissions <0666> expected <0600> actual
ERROR: /devices/pseudo/cvc@0:cvc
     pathname does not exist
ERROR: /devices/pseudo/cvcredir@0:cvcredir
      pathname does not exist

For Solaris Flash installations, errors similar to the previous output are also displayed for the SUNWcsr package.

Workaround: Choose one of the following workarounds.

Workaround 1: Run the pkgchk -nf command on the SUNWcsd Core Solaris Devices package. This workaround applies if you installed Solaris 10 OS from CD or DVD media, or from a network installation image.


# pkgchk -nf SUNWcsd

Workaround 2: This workaround applies to clone systems on which you want to install the Solaris 10 OS by using a Solaris Flash archive. Perform the following steps before you install the archive:

  1. On the master system, create a /etc/flash/reboot directory:


    # mkdir -p /etc/flash/reboot #

  2. On the master system, create the following script in a text editor.


    #! /usr/bin/sh 
# echo " Doing pkgchk -nf SUNWcsd" 
pkgchk -nf SUNWcsd 
echo " Doing pkgchk -nf SUNWcsr" 
pkgchk -nf SUNWcsr 
exit 0

  3. Save the script with the file name pkgchk.cleanup in the /etc/flash/reboot directory.

  4. Ensure that the pkgchk.cleanup script is executable.


    # cd /etc/flash/reboot
# chmod a+x pkgchk.cleanup

  5. Create the Solaris Flash archive.

  6. Install the archive on the clone system.

However, if you have already installed the Solaris 10 OS with a Solaris Flash archive, then perform the following steps to resolve the error.

  1. Run the pkgchk -nf command on the SUNWcsd Core Solaris Devices package.


    # pkgchk -nf SUNWcsd

  2. Run the pkgchk -nf command on the SUNWcsr Core Solaris (Root) package on the clone system.


    # pkgchk -nf SUNWcsr

SPARC: Error Messages Might Be Seen When Installing Solaris By Using a Network Image (4704046)

You can install Solaris by using a network image. If that image is created by combining the Solaris 10 Software CDs, the following error messages might be seen:


Error message:    Jun 18 10:51:53 b26b statd[149]: [ID 514559 
daemon.error] svc_tp_create: Could not register prog 100024 vers 1 on udp 
Jun 18 10:51:53 b26b statd[149]: [ID 514559 daemon.error] svc_tp_create: 
Could not register prog 100024 vers 1 on tcp 
Jun 18 10:51:53 b26b statd[149]: [ID 514559 daemon.error] svc_tp_create: 
Could not register prog 100024 vers 1 on ticlts 
Jun 18 10:51:53 b26b statd[149]: [ID 514559 daemon.error] svc_tp_create: 
Could not register prog 100024 vers 1 on ticotsord 
Jun 18 10:51:53 b26b statd[149]: [ID 514559 daemon.error] svc_tp_create: 
Could not register prog 100024 vers 1 on ticots 
Jun 18 10:51:53 b26b statd[149]: [ID 447212 daemon.error] statd: 
unable to create (SM_PROG, SM_VERS) for netpath.

Workaround: None. Ignore the message. After the system boots, the Network File System (NFS) functions normally.

SPARC: Systems With Multiple Interfaces Recognize All Interfaces as Usable After Installation or Upgrade (4640568)

If you install or upgrade to the Solaris 10 release on a system with multiple network interfaces, the system recognizes all system interfaces as usable. Interfaces that are not connected to the network, or that are not intended for use, appear in the output of the ifconfig -a command. Additionally, interfaces with identical Ethernet addresses might be assigned identical IP addresses. The following error message is displayed:


ifconfig: setifflags: SIOCSLIFFLAGS: qfe3: Cannot assign requested address

This problem also occurs on systems that have the local-mac-address PROM variable set to false. The problem occurs because all interfaces are configured with the same IP address.

Workaround: Choose one of the following workarounds:

  • To plumb only the configured interface, reboot the system after the initial boot.

  • To assign a different IP address to each network interface, set the local-mac-address PROM variable to true in one of the following ways:

    • At the ok prompt, type the following command:


      ok setenv local-mac-address? true

    • In a terminal window, type the following command as superuser:


      # eeprom local-mac-address?=true

Warnings Might Occur When a File System Is Created (4189127)

When a file system is created during installation, one of the following warning messages might be displayed:


Warning: inode blocks/cyl group (87) >= data blocks (63) in last
cylinder group. This implies 1008 sector(s) cannot be allocated.

Or:


Warning: 1 sector(s) in last cylinder unallocated

The warning occurs when the size of the file system that you created does not equal the space on the disk that is being used. This discrepancy can result in unused space on the disk that is not incorporated into the indicated file system. This unused space is not available for use by other file systems.

Workaround: Ignore the warning message.

Upgrade Issues and Bugs

Note –

For the latest information about upgrade support beginning with the Solaris 10 1/06 release, see Changes in Upgrade Support for Solaris Releases.

This section describes upgrade bugs. Some might occur while you are upgrading to Solaris 10 OS. Others might occur after you have completed upgrading.

Device ID Discrepancies After Upgrading From Solaris 9 9/04 OS

In this Solaris 10 release, Solaris Volume Manager displays device ID output in a new format. Solaris 9 9/04 OS, which introduced device ID support in disk sets, does not recognize the new format. When you upgrade to Solaris 10 OS from the Solaris 9 9/04 release, device IDs that are associated with existing disk sets are not updated in the Solaris Volume Manager configuration. If you need to revert to Solaris 9 9/04 OS, configuration changes made to disk sets after the upgrade might not be available to Solaris 9 9/04 OS. For more information, see Chapter 25, Troubleshooting Solaris Volume Manager (Tasks), in Solaris Volume Manager Administration Guide.

SPARC: Upgrading From Solaris 9 Releases With Recommended Patch Cluster Partially Succeeds (6202868)

For a system that is running a Solaris 9 release with the recommended patch cluster installed, upgrading to Solaris 10 OS only partially succeeds. This problem affects systems that are running the following releases with the Solaris 9 recommended patch cluster installed.

  • Solaris 9 release

  • Solaris 9 9/02 release

  • Solaris 9 12/02 release

  • Solaris 9 4/03 release

  • Solaris 9 8/03 release

  • Solaris 9 12/03 release

  • Solaris 9 4/04 release

When you upgrade to the Solaris 10 release, the SUNWcti2x package is not successfully removed from the system.

Workaround: Choose one of the following workarounds.

  • To avoid this problem, apply patch ID 117426-03, or a later version, to the system before you upgrade to the Solaris 10 OS.

    To download this patch, go to http://sunsolve.sun.com.

  • If you encounter this problem during the upgrade, follow these steps.

    1. In a text editor, comment out the following line in the /var/sadm/pkg/SUNWcti2x/install/preremove file.


      rem_drv -b ${BASEDIR} sc_nct || EXIT=1

    2. Remove the SUNWcti2x package.


      # pkgrm SUNWcti2x

Obsolete Uninstallers Not Removed When You Use Solaris Live Upgrade to Upgrade From Previous Solaris Releases (6198380)

If you use Solaris Live Upgrade to upgrade from the Solaris 8 or Solaris 9 releases to Solaris 10 OS, obsolete uninstaller programs are not removed. These uninstaller programs from the previous OS remain in the system's /var/sadm/prod directory.

The following obsolete uninstallers are not removed.


uninstall_Alternate_Pathing_2_3_1.class
uninstall_CDRW_1_1.class o uninstall_CDRW_1_0.class
uninstall_Bonus_Localization_-_Catalan_CDE_Desktop.class
uninstall_Bonus_Localization_-_Polish_CDE_Desktop.class
uninstall_Bonus_Localizations_-_Russian_CDE_Desktop.class
uninstall_Capacity_on_Demand_1_0.class
uninstall_Java3D_1_3_1.class
uninstall_Java3D_1_3.class
uninstall_Java3D_1_2_1_04.class
uninstall_Java3D_1_2_1_03.class
uninstall_Lights_Out_Management_2_0.class
uninstall_Man_Page_Supplement.class
uninstall_OpenGL_1_3.class
uninstall_OpenGL_1_2_3.class
uninstall_Netra_ct_Platform_1_0.class
uninstall_Netra_t11xx_Alarms_2_0.class
uninstall_Netscape_6_2_3.class
uninstall_Netscape_6_2_1_Beta.class
uninstall_PC_launcher_1_0_2.class
uninstall_PC_launcher_1_0_1_PCfileviewer_1_0_1.class
uninstall_RSC_2_2_2.class
uninstall_RSC_2_2_1.class
uninstall_RSC_2_2.class
uninstall_ShowMeTV_1_3.class
uninstall_Solaris_9_French_Localization.class
uninstall_Solaris_9_German_Localization.class
uninstall_Solaris_9_Hong_Kong_Traditional_Chinese_Localization.class
uninstall_Solaris_9_Italian_Localization.class
uninstall_Solaris_9_Japanese_Localization.class
uninstall_Solaris_9_Korean_Localization.class
uninstall_Solaris_9_Simplified_Chinese_Localization.class
uninstall_Solaris_9_Spanish_Localization.class
uninstall_Solaris_9_Swedish_Localization.class
uninstall_Solaris_9_Traditional_Chinese_Localization.class
uninstall_Solaris_On_Sun_Hardware_Documentation.class
uninstall_Sun_Hardware_AnswerBook.class
uninstall_SunATM_5_0.class
uninstall_SunATM_5_1.class
uninstall_SunFDDI_PCI_3_0.class
uninstall_SunFDDI_SBus_7_0.class
uninstall_Sun_Fire_880_FC-AL_Backplane_Firmware_1_0.class
uninstall_Sun_Fire_B10n_Load_Balancing_Blade_1_1.class
uninstall_SunForum_3_1.class
uninstall_SunForum_3_2.class
uninstall_SunHSI_PCI_3_0.class
uninstall_SunHSI_SBus_3_0.class
uninstall_SunScreen_3_2.class
uninstall_SunVTS_5_1_PS6.class
uninstall_SunVTS_5_1_PS5.class
uninstall_SunVTS_5_1_PS4.class
uninstall_SunVTS_5_1_PS3.class
uninstall_SunVTS_5_1_PS2.class
uninstall_SunVTS_5_1_PS1.class
uninstall_SunVTS_5_0.class
uninstall_System_Management_Services_1_4.class
uninstall_System_Management_Services_1_3.class
uninstall_System_Management_Services_1_2.class
uninstall_System_Service_Processor_3_5.class
uninstall_WBEM_DR_1_0.class
uninstall_Web_Start_Wizards_SDK_3_0_2.class
uninstall_Web_Start_Wizards_SDK_3_0_1.class
uninstall_Web_Start_Wizards_SDK.class
uninstall_XML_Libraries_2_4_12.class

Workaround: After you upgrade the system, manually remove the obsolete uninstallers in the /var/sadm/prod directory.

Configuration File pam.conf Not Automatically Updated After an Upgrade (5060721)

This Solaris 10 release introduces changes in pam_ldap functionality. When you upgrade to the current release, pam_ldap configurations in your existing pam.conf configuration file are not updated to reflect these changes. If pam_ldap configuration is detected, the CLEANUP file that is generated at the end of the upgrade contains the following notification:


/etc/pam.conf please examine/update the pam_ldap configuration 
because its functionality has changed, 
refer to pam_ldap(5) documentation for more information

Workaround: After the upgrade, examine /etc/pam.conf. If necessary, modify this file manually to be compatible with the new functionalities of pam_ldap. The modifications involve password prompting such as the use_first_pass and try_first_pass options as well as password updates. For more information about updating pam.conf, refer to the pam_ldap(5) man page and documentation.

Solstice DiskSuite Configurations Not Converted to Solaris Volume Manager Format When You Upgrade With Solaris Live Upgrade (4915974)

If you upgrade from the Solaris 8 release to the Solaris 10 release with Solaris Live Upgrade, an error might occur. Your Solstice DiskSuiteTM 4.2.1 configurations might not be converted to Solaris Volume Manager configurations. All RAID-0, RAID-1, and other volumes such as mirrors, stripes, and submirrors might be lost.

The problem occurs if you have applied patches in a particular order to the Solaris 8 system before you upgrade. To determine if your system might experience the problem, check the version information of the SUNWmdr package on the Solaris 8 system. Type the following command:


# grep VERSION /var/sadm/pkg/SUNWmdr/pkginfo

Your system might be affected by the problem if the command generates the following output:


PKG_CLIENT_VERSION=8
VERSION=4.2.1,REV=1999.12.03.10.00

Workaround: Follow these steps.

  1. Upgrade the alternate boot environment by using Solaris Live Upgrade.

    See Solaris 10 Installation Guide: Solaris Live Upgrade and Upgrade Planning for more information.

  2. Mount the upgraded boot environment disk, but do not activate the upgraded boot environment.


    # mnt -F ufs /mnt/c0t0d0slice_number /mnt

    In the previous example, slice_number refers to the slice that contains the upgraded boot environment.

  3. Check the /etc/system file for MDD Database information entries, such as the following text:


    * Begin MDD database info (do not edit)
set md:mddb_bootlist1="sd:7:16 sd:7:1050 sd:7:2084"
* End MDD database info (do not edit)

  4. In a text editor, open the /kernel/drv/md.conf file for the upgraded boot environment.


    # cd /kernel/drv
# vi md.conf

  5. Modify md.conf as follows:

    1. Add the mddb_bootlist information from the /etc/system file that you checked in Step 3. Make sure that you also include the commented “Begin” and “End” lines that precede and follow the information.

    2. For each sd disk entry on the mddb_bootlist line, add the text string :id0 at the end of the disk entry.

    3. Below the mddb_bootlist entry, insert the following new line:


      md_devid_destroy=1;

    The following example shows how the md.conf file might be modified:


    # Begin MDD database info (do not edit)
mddb_bootlist1="sd:7:16:id0 sd:7:1050:id0 sd:7:2084:id0";
md_devid_destroy=1;
# End MDD database info (do not edit)

  6. Save the md.conf file.

  7. Delete the mddb_bootlist entry from the /etc/system file.

  8. Unmount the upgraded boot environment disk.

  9. Activate the upgraded boot environment.

Invalid Error Messages Are Displayed When You Use Solaris Live Upgrade to Upgrade From the Solaris 7 Release (4872151)

If you use Solaris Live Upgrade to upgrade from the Solaris 7 release to the Solaris 10 OS, error messages similar to the following are displayed.


Removal of package was successful.
/a//var/sadm/system/admin/upgrade_script: /bin/prodreg: not found
/a//var/sadm/system/admin/upgrade_script: /bin/prodreg: not found
/a//var/sadm/system/admin/upgrade_script: /bin/prodreg: not found

This error occurs because the Solaris Product Registry (prodreg) software is not included in the Solaris 7 release.

Workaround: Ignore the error message. These errors do not affect the installation.

Installer Text Display Problem When Using Solaris Live Upgrade (4736488)

When using the Solaris Live Upgrade luupgrade(1M) command with the -i option to complete an upgrade of an inactive boot environment, the text that the installers display might be unreadable in some languages. The text is corrupted when the installers request fonts that do not exist on the older release that is on the current boot environment.

Workaround: Choose one of the following workarounds:

  • Use a combined network installation image to do the installation.

  • Enable the C locale by setting the environment variable on your system.

    • If you are using the Bourne shell or Korn shell, follow these steps:

      1. Set the C locale.


        # LANG=C; export LANG

      2. Begin the installation.

    • If you are using the C shell, follow these steps:

      1. Type the following:


        # csh

      2. Set the C locale.


        # setenv LANG C

      3. Begin the installation.

SPARC: Removal of SUNWjxcft Package Records Error During Upgrade (4525236)

When you upgrade from the Solaris 8 software to the Solaris 10 release, a problem is encountered when the SUNWjxcft package is removed. The following error message is recorded in the upgrade_log file:


Removing package SUNWjxcft: 
Can't open /a/usr/openwin/lib/locale/ja/X11/fonts/TTbitmaps/fonts.upr 
Can't open /a/usr/openwin/lib/locale/ja/X11/fonts/TTbitmaps/fonts.scale 
Can't open /a/usr/openwin/lib/locale/ja/X11/fonts/TTbitmaps/fonts.alias 
Can't open /a/usr/openwin/lib/locale/ja/X11/fonts/TT/fonts.upr 
Can't open /a/usr/openwin/lib/locale/ja/X11/fonts/TT/fonts.scale 
Can't open /a/usr/openwin/lib/locale/ja/X11/fonts/TT/fonts.alias 
Removal of <SUNWjxcft> was successful

Workaround: Ignore the error message.

Upgrading to Solaris 10 Release Might Disable Existing Secure Shell Daemon (sshd) (4626093)

If you upgrade to the Solaris 10 release on a system that is running a third-party Secure Shell, such as OpenSSH from the /etc/init.d/sshd daemon, the upgrade disables the existing Secure Shell daemon. During an upgrade, Solaris 10 software overwrites the contents of /etc/init.d/sshd.

Workaround: Choose one of the following workarounds:

  • If you do not want the Secure Shell protocol server program on your system, do not install the SUNWsshdr and SUNWsshdu packages during the upgrade.

  • If you do not want the Secure Shell protocol server or client programs on your system, do not install the Secure Shell Cluster (SUNWCssh) during the upgrade.

Upgrade Fails if /export Directory Is Near Capacity (4409601)

If the /export directory is near full capacity when you upgrade to the Solaris 10 release, space requirements for /export are miscalculated. The upgrade then fails. This problem commonly occurs if a diskless client is installed. Another instance of when the problem occurs is when third-party software is installed in the /export directory. The following message is displayed:


WARNING: Insufficient space for the upgrade.

Workaround: Before you upgrade, choose one of the following workarounds:

  • Rename the /export directory temporarily, until the upgrade is completed.

  • Temporarily comment out the /export line in the /etc/vfstab file until the upgrade is completed.

  • If /export is a separate file system, then unmount /export before you perform the upgrade.

Upgrading Diskless Client Servers and Clients (4363078)

If your system currently supports diskless clients that were installed with the Solstice AdminSuiteTM 2.3 Diskless Client tool, you must perform the following two steps:

  1. Delete all existing diskless clients that are the same Solaris version and architecture as the server.

  2. Install or upgrade to the Solaris 10 release.

For specific instructions, see the System Administration Guide: Basic Administration.

If you attempt to install the Solaris 10 software over existing diskless clients, the following error message might be displayed:


The Solaris Version (Solaris version-number) on slice 
<xxxxxxxx> cannot 
be upgraded. 
There is an unknown problem with the software configuration installed 
on this disk.

In this error message, version-number refers to the Solaris version that is currently running on your system. <xxxxxxxx> refers to the slice that is running this version of the Solaris software.

Additional Installation Issues

This section describes issues that are related to the installation of the Solaris OS.

StarOffice and StarSuite Software Cannot Coexist in the Same System

When you install Solaris 10 OS, either the StarOffice or the StarSuiteTM software is also automatically installed, depending on which language you select. The languages and the corresponding software that these languages support are listed as follows:

Selected Language 

Supported Software 

Chinese, Japanese, Korean 

StarSuite 

Other languages 

StarOffice 

StarOffice and StarSuite cannot coexist in the same system. If you want to replace a software that you accidentally installed, follow these steps.

  1. Insert the Solaris 10 Software - 3 CD or Solaris 10 Operating System DVD in the drive.

  2. Become superuser.

  3. Change to the Product directory, for example, /cdrom/cdrom0/Solaris_10/Product.

  4. Replace the software.

    • To replace StarOffice with StarSuite, use the following commands:


      # pkgrm SUNWsogm SUNWsom
# pkgadd -d . SUNWsoagm SUNWsoam

    • To replace StarSuite with StarOffice, use the following commands:


      # pkgrm SUNWsoagm SUNWsoam
# pkgadd -d . SUNWsogm SUNWsom

Cannot Install Documentation Packages With Names Longer Than Nine Characters on Documentation Servers Running Solaris 7 or Solaris 8 Software

Some localized documentation collections in PDF format have package names that are longer than nine characters. To install these PDF collections on servers that are running Solaris 7 or 8 software, you must first install two patches.

Workaround: For instructions on how to install these patches, see the Solaris Documentation Important Information file on Solaris 10 OS Documentation DVD. This file is located in the following directory:


mount-point/README/locale/install_
locale.html

Additional Related Locales Might Be Installed

When you select a locale for your installation, additional related locales might also be installed. This change in behavior occurs in the Solaris 10 release because all full locales, with message translations, and the Asian and Japanese partial locales, locale enabler, have been repackaged based on language support for locales. Other partial locales are still packaged and installed based on geographic region, such as Central Europe.

Languages CD Installs All Languages By Default With Solaris Live Upgrade (4898832)

If you use Solaris Live Upgrade with multiple CDs to install the Solaris 10 release, the Languages CD installs all languages by default.

After the installation, if you log in to the system in a locale that is different than the locale you selected during installation, garbled characters might be displayed. After you log in to any of these locales, the English locale is displayed.

Workaround: During installation, select the custom install option. Uncheck any languages that you do not want to install during the Languages CD installation.