C H A P T E R 3 |
Open Issues |
This chapter discusses:
If you have SunScreen SKIP 1.1.1 software currently installed on your system, you should remove the SKIP packages before installing or upgrading to the
Solaris 9 9/05 operating environment. Remove these packages: SICGbdcdr, SICGc3des, SICGcdes, SICGcrc2, SICGcrc4, SICGcsafe, SICGes, SICGkdsup, SICGkeymg, SICGkisup
The SunVTS diagnostic software now has its own release note document. For the latest information on features and bugs, see the SunVTS 5.1 Patch Set 10 Release Notes, part number 819-2825-10, at:
This information on Sun Remote System Control (RSC) 2.2.3 hardware and software issues can be found in the Sun Remote System Control (RSC) 2.2.3 Release Notes, part number 819-2022, at:
SunForum video cannot be used in an environment that supports only 8 bits per pixel. In particular, video does not work when used on PGX8, PGX24, and PGX64 in defdepth 8 mode.
To install the Solaris 9 9/05 operating environment on specific Sun Fire and Netra servers, you must first update the firmware on the server. If you do not update the firmware before you install the Solaris 9 9/05 software, the server panics. This problem affects the following servers:
If this problem occurs, the following error message is displayed:
Workaround: Update the firmware by applying the appropriate firmware patch.
For Sun Fire 3800, 4800, 4810 or 6800 servers, apply one of the following patches:
For Sun Fire V1280 or Netra 1280 servers, apply Patch ID 113751-03 (5.13.0013 firmware update).
Firmware update patches are available at:
Be sure to download and apply the most current version of the patch.
Some X1 systems require a PROM patch from Solaris 8 before running Solaris 9. When booting a Netra X1 in the Solaris 9 OE, the following error message might appear:
Check the system's PROM version.
Install patch 111092-02 or later (this is a prerequisite to Step 3). Do not reboot the system between Steps 2 and 3.
Install patch 111952-02 or later.
This problem only applies to Netra X1 systems, not to Sun Fire V100 systems.
Changing a target ID on a running Sun StorEdge T3 results in a missing LUN.
Workaround: Issue a LIP again using the vol unmount and vol mount commands. The host now sees the device.
A Sun StorEdge T3 system might panic if an application uses the HTTP interface to send tokens with out-of-range parameters.
Under repeated stress using PCI hotplug, RCM fails with error code 7 on a Sun Fire V880 system.
A Sun Fire V880 might not automatically recognize the XVR-4000 graphics accelerator after a first-time fresh Solaris installation, resulting in a blank screen on monitors connected to the XVR-4000. This problem can also occur when re-installing a version of Solaris that does support the XVR-4000 (Solaris 9 4/03 or later, or other Solaris versions supported by a special installation DVD included with the XVR-4000), replacing a version of Solaris that does not support the XVR-4000.
Begin your Solaris installation using either a ttya or PCI-based graphics card with the console.
Install Solaris software as usual.
Perform a power cycle.
The XVR-4000 should be recognized now as the new default screen.
The fix for Bug ID 4717004 removed the bbc and gptwo drivers. Even though these drivers have been deleted, the V880 software delivered with Solaris 9 9/05 still attempts to load these drivers, causing the following messages to appear during boot:
Workaround: These messages are benign and can safely be ignored.
The SUNWCXall software metacluster must be installed on a Sun Fire V250.
Running the trapstat command on a Sun Fire 6800 might result in a crash and error messages naming fatal conditions.
On a Sun Fire 4800 system, the command psrinfo -p might not return the correct number of CPUs. The command psrinfo -sp causes a segmentation fault. This problem might be masked if an UltraSPARC IV board is part of the system.
Workaround: use the command options psrinfo and psrinfo -s instead.
If an UltraSPARC IV+ board is added to a Sun Fire V1280 domain via dynamic reconfiguration, the prtconf and prtdiag programs do not report the nodes corresponding to the UltraSPARC IV+ CPUs as being in the system. These reports remain unchanged until the system is rebooted.
For domains containing CPU/MCPU boards flashed at LPOST level 5.13.3 or lower, an LPOST error might cause a failure to boot the Solaris OE and/or cause a system to hang.Workaround: SMS 1.2 patch 112829-05 (or higher) contains an updated LPOST flash image. This patch is available at:
http://sunsolve.sun.com
This bug is fixed in SMS 1.3. Upgrading to SMS 1.2 or later, from previous versions, is recommended.To find out the LPOST level flashed on your system's CPU/MCPU boards, type:
Where X is the letter (A-R) of the domain.
To find out whether the patch already exists on your system, type:
If Patch 112829-05 is not installed, no data is returned. If this happens, perform the following procedure:
Apply patch 112829-05 to both system controllers according to the instructions in the patch README file.
Use the flashupdate command to update the LPOST image on CPU/MCPU boards.
Refer to the flashupdate man page for specific command syntax.
On Sun Fire 15K/12K systems, SMS might return the unum string with an incorrect value. The string is in the form:
Where w is 0 - 17, x is 0 - 3, y is 0 - 1, and z is from one to four digits long. (All ranges are inclusive.) This bug might cause the P# number to be 0 - 7 instead of 0 - 3.
Workaround: If the P# is a value between 4 and 7, subtract 4 to get the correct value.
This bug affects Sun Blade 1000 or 2000 systems that have the XVR-1000 graphics accelerator installed. If Xsun dies or exits while the graphics bus is under power management, the system panics.
Workaround: Add the following line to the system's /etc/power.conf file:
The Solaris Installation CD automatically installs the network drivers from the Supplement CD. Drivers for the SunATM and SunHSI/P cards are now installed by default when the Solaris 9 9/05 operating environment is installed. Error messages might be displayed if you do not have some of the corresponding hardware installed. You can ignore these error messages.
A failure in spec_open causes mount failure of points referenced in the vfstab entry.
Workaround: Type the following command:
This loads the ssd driver and attaches all device instances.
A dynamic reconfiguration operation on the last connection to a multipathed device can be detached without warning.
The device driver does not detect when the cable is disconnected from the port and pulled away from the connector edge of the dual-channel FC PCI card.
Workaround: Install Patch 111097-08 or later. Further information about this patch is available at the SunSolve web site at: http://sunsolve.sun.com
Pulling the fibre cable going to controller A0 on an A5000 causes any subsequent DR operation for detaching the system board to fail. Reconnecting the fibre cable does not enable a successful DR detach operation.
Power management of qlc causes the kernel to panic on an ASSERTION failure in the qlc driver code.
Workaround: Set the following in /etc/system to avoid the assertion:
qlc might remain offline, preventing devices appearing after boot.
Workaround: Generate a LIP on the link to get the port in an Online state. You can generate LIP on the HBA port by running the command luxadm -e forcelip.
A system using the kadb command to debug a live system can go into a repeating loop of incomplete error messages when the OpenBoot PROM's master CPU is changed. While a reset will restore the system to operation, the traces of the original failure are lost, resulting in an inability to perform the attempted diagnosis of a fatal reset.
Upgrade to the latest version of OpenBoot PROM.
Before switching, raise pil to f with the following command:
The Sun StorEdge network foundation software driver does not honor untagged commands, and allows overlapped untagged commands.
The default timeout value for the SCSI portion of the SunSwift PCI Ethernet/SCSI host adapter (X1032A) card does not meet the timeout requirements of Sun's SCSI DVD-ROM drive (X6168A). With marginal media, the DVD-ROM occasionally experiences timeout errors. The only exceptions are Sun Fire 6800, 4810, 4800, and 3800 systems, which overwrite the SCSI timeout value by means of OpenBoot PROM.
Workaround for other platforms: Use the on-board SCSI interfaces or DVD-ROM compatible SCSI adapters, such as X1018A (SBus: F501-2739-xx) or X6540A (PCI: F375-0005-xx).
When MPxIO is enabled during a Dynamic Reconfiguration operation, the qlc driver might suspend and cause a domain panic.
cPCI and dual-channel FC PCI cards might encounter a SCSI timeout or "giving up" error message due to a firmware problem.
Installation of the Solaris operating environment on drives with 96 GBytes or more might fail using some installation methods.
Workaround: Begin Solaris installation using the Solaris 9 9/05 Software 1 of 2 CD.
The OpenBoot PROM command probe-scsi or probe-scsi-all must be followed by the reset-all command on sun4u systems before booting. Otherwise, a subsequent boot or boot -v might hang.
The package SUNWfcbpl has been removed from the Supplement CD for this Solaris release to prevent installation problems on some systems. This package is no longer needed to upgrade Fibre Channel backplane firmware. Download Patch 117814 if an upgrade of Fibre Channel backplane firmware to version 922A or later is needed.
Copyright © 2005, Sun Microsystems, Inc. All Rights Reserved.