This software and related documentation are provided under a license agreement containing restrictions on use and disclosure and are protected by intellectual property laws. Except
as expressly permitted in your license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify, license, transmit, distribute, exhibit, perform,
publish, or display any part, in any form, or by any means. Reverse engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is
prohibited.
The information contained herein is subject to change without notice and is not warranted to be error-free. If you find any errors, please report them to us in writing.
If this is software or related documentation that is delivered to the U.S. Government or anyone licensing it on behalf of the U.S. Government, then the following notice is applicable:
U.S. GOVERNMENT END USERS: Oracle programs, including any operating system, integrated software, any programs installed on the hardware, and/or documentation,
delivered to U.S. Government end users are "commercial computer software" pursuant to the applicable Federal Acquisition Regulation and agency-specific supplemental
regulations. As such, use, duplication, disclosure, modification, and adaptation of the programs, including any operating system, integrated software, any programs installed on the
hardware, and/or documentation, shall be subject to license terms and license restrictions applicable to the programs. No other rights are granted to the U.S. Government.
This software or hardware is developed for general use in a variety of information management applications. It is not developed or intended for use in any inherently dangerous
applications, including applications that may create a risk of personal injury. If you use this software or hardware in dangerous applications, then you shall be responsible to take all
appropriate fail-safe, backup, redundancy, and other measures to ensure its safe use. Oracle Corporation and its affiliates disclaim any liability for any damages caused by use of this
software or hardware in dangerous applications.
Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of their respective owners.
Intel and Intel Xeon are trademarks or registered trademarks of Intel Corporation. All SPARC trademarks are used under license and are trademarks or registered trademarks of
SPARC International, Inc. AMD, Opteron, the AMD logo, and the AMD Opteron logo are trademarks or registered trademarks of Advanced Micro Devices. UNIX is a registered
trademark of The Open Group.
This software or hardware and documentation may provide access to or information about content, products, and services from third parties. Oracle Corporation and its affiliates are
not responsible for and expressly disclaim all warranties of any kind with respect to third-party content, products, and services unless otherwise set forth in an applicable agreement
between you and Oracle. Oracle Corporation and its affiliates will not be responsible for any loss, costs, or damages incurred due to your access to or use of third-party content,
products, or services, except as set forth in an applicable agreement between you and Oracle.
Access to Oracle Support
Oracle customers that have purchased support have access to electronic support through My Oracle Support. For information, visit http://www.oracle.com/pls/topic/lookup?
ctx=acc&id=info or visit http://www.oracle.com/pls/topic/lookup?ctx=acc&id=trs if you are hearing impaired.
Ce logiciel et la documentation qui l'accompagne sont protégés par les lois sur la propriété intellectuelle. Ils sont concédés sous licence et soumis à des restrictions d'utilisation et
de divulgation. Sauf stipulation expresse de votre contrat de licence ou de la loi, vous ne pouvez pas copier, reproduire, traduire, diffuser, modifier, accorder de licence, transmettre,
distribuer, exposer, exécuter, publier ou afficher le logiciel, même partiellement, sous quelque forme et par quelque procédé que ce soit. Par ailleurs, il est interdit de procéder à toute
ingénierie inverse du logiciel, de le désassembler ou de le décompiler, excepté à des fins d'interopérabilité avec des logiciels tiers ou tel que prescrit par la loi.
Les informations fournies dans ce document sont susceptibles de modification sans préavis. Par ailleurs, Oracle Corporation ne garantit pas qu'elles soient exemptes d'erreurs et vous
invite, le cas échéant, à lui en faire part par écrit.
Si ce logiciel, ou la documentation qui l'accompagne, est livré sous licence au Gouvernement des Etats-Unis, ou à quiconque qui aurait souscrit la licence de ce logiciel pour le
compte du Gouvernement des Etats-Unis, la notice suivante s'applique :
U.S. GOVERNMENT END USERS: Oracle programs, including any operating system, integrated software, any programs installed on the hardware, and/or documentation,
delivered to U.S. Government end users are "commercial computer software" pursuant to the applicable Federal Acquisition Regulation and agency-specific supplemental
regulations. As such, use, duplication, disclosure, modification, and adaptation of the programs, including any operating system, integrated software, any programs installed on the
hardware, and/or documentation, shall be subject to license terms and license restrictions applicable to the programs. No other rights are granted to the U.S. Government.
Ce logiciel ou matériel a été développé pour un usage général dans le cadre d'applications de gestion des informations. Ce logiciel ou matériel n'est pas conçu ni n'est destiné à être
utilisé dans des applications à risque, notamment dans des applications pouvant causer un risque de dommages corporels. Si vous utilisez ce logiciel ou ce matériel dans le cadre
d'applications dangereuses, il est de votre responsabilité de prendre toutes les mesures de secours, de sauvegarde, de redondance et autres mesures nécessaires à son utilisation dans
des conditions optimales de sécurité. Oracle Corporation et ses affiliés déclinent toute responsabilité quant aux dommages causés par l'utilisation de ce logiciel ou matériel pour des
applications dangereuses.
Oracle et Java sont des marques déposées d'Oracle Corporation et/ou de ses affiliés. Tout autre nom mentionné peut correspondre à des marques appartenant à d'autres propriétaires
qu'Oracle.
Intel et Intel Xeon sont des marques ou des marques déposées d'Intel Corporation. Toutes les marques SPARC sont utilisées sous licence et sont des marques ou des marques
déposées de SPARC International, Inc. AMD, Opteron, le logo AMD et le logo AMD Opteron sont des marques ou des marques déposées d'Advanced Micro Devices. UNIX est une
marque déposée de The Open Group.
Ce logiciel ou matériel et la documentation qui l'accompagne peuvent fournir des informations ou des liens donnant accès à des contenus, des produits et des services émanant de
tiers. Oracle Corporation et ses affiliés déclinent toute responsabilité ou garantie expresse quant aux contenus, produits ou services émanant de tiers, sauf mention contraire stipulée
dans un contrat entre vous et Oracle. En aucun cas, Oracle Corporation et ses affiliés ne sauraient être tenus pour responsables des pertes subies, des coûts occasionnés ou des
dommages causés par l'accès à des contenus, produits ou services tiers, ou à leur utilisation, sauf mention contraire stipulée dans un contrat entre vous et Oracle.
Accès aux services de support Oracle
Les clients Oracle qui ont souscrit un contrat de support ont accès au support électronique via My Oracle Support. Pour plus d'informations, visitez le site http://www.oracle.com/
pls/topic/lookup?ctx=acc&id=info ou le site http://www.oracle.com/pls/topic/lookup?ctx=acc&id=trs si vous êtes malentendant.
Page 5
Contents
Using This Documentation ............. ................ ................ ................ ................ ... 11
Linux Issues ...................................................................................................... 51
Oracle and Red Hat Linux 7.2 Installation Fails With AER Error on Systems With
Four or More Sun InfiniBand Dual Port 4x QDR PCIe Low Profile Host Channel
Hotplugging PCIe EM Card Fails When Replacing the Original Card With One of a
Different Type (CR 7003634) .......................................................................... 90
Messages Complain of Failure to Allocate I/O Resources (CR 6984329) .................. 90
DIMM Failure Causes Other DIMMs To Be Disabled (CR 6929978) ....... ............... 91
10Sun Fire X4800 Server Product Notes • July 2017
Page 11
Using This Documentation
■
Overview – Provides late-breaking information, issue status, and product announcements
for Oracle's Sun Fire X4800 Server.
■
Audience – System administrators, network administrators, and service technicians.
■
Required knowledge – Advanced understanding of server systems.
Product Documentation Library
Documentation and resources for this product and related products are available at http://
docs.oracle.com/cd/E19140-01/index.html.
Feedback
Provide feedback about this documentation at http://www.oracle.com/goto/docfeedback.
Using This Documentation11
Page 12
12Sun Fire X4800 Server Product Notes • July 2017
Page 13
Important Operating Notes
This section provides information about critical issues that affect your server. These include:
■
“Server Security, Software Releases, and Critical Patch Updates” on page 13
■
“IMPORTANT - Install Latest OS Updates, Patches, and Firmware” on page 14
■
“Diagnosing SAS Data Path Failures on Servers Using MegaRAID Disk
Controllers” on page 15
■
“Oracle ILOM Deprecation Notice for IPMI 2.0 Management Service Present
Behavior” on page 16
■
“Oracle ILOM Deprecation Notice for Default Self-Signed Certificate” on page 17
■
“Broken Links in Sun Server X4800 Documentation Library” on page 17
For known issues affecting system components, see:
■
“Hardware Issues” on page 31
■
“Oracle Solaris 10 Operating System Issues” on page 43
■
“Linux Issues” on page 51
■
“Oracle VM Issues” on page 61
■
“Windows Operating System Issues” on page 63
■
“ESX Issues” on page 67
■
“Oracle ILOM Enhancements and Issues” on page 71
■
“Oracle Hardware Installation Assistant Issues” on page 87
■
“BIOS Issues” on page 89
Server Security, Software Releases, and Critical Patch
Updates
To ensure continued security of your system, Oracle strongly recommends that you apply the
latest Software Releases. Server Software Releases include Oracle ILOM, BIOS, and other
firmware updates, often referred to as “patches.” Oracle publishes these patches regularly on
the My Oracle Support site. Applying these patches helps ensure optimal system performance,
security, and stability. You can identify the latest Software Release for your system at: http://
Important Operating Notes13
Page 14
IMPORTANT - Install Latest OS Updates, Patches, and Firmware
www.oracle.com/technetwork/systems/patches/firmware/release-history-jsp-138416.
html
To download a Software Release, go to My Oracle Support at: https://support.oracle.com
Oracle notifies customers about security vulnerability fixes for all its products four times
a year through the Critical Patch Update (CPU) program. Customers should review the
CPU advisories to ensure that the latest software release updates are applied to their Oracle
products. Note that updates for Engineered Systems are specifically published for a specific
Engineered Systems product (that is, you need not look at specific updates for individual
software components included in your Engineered System). For more information about the
Oracle CPU program, go to: http://www.oracle.com/technetwork/topics/security/
alerts-086861.html
Oracle also recommends that you update to the latest operating system release when it becomes
available. Although a minimum operating system release is supported, updating to the latest
OS release ensures that you have the most up-to-date software and security patches. To confirm
that you have the latest OS release, refer to the Oracle Hardware Compatibility Lists. See
“Supported Operating Systems” on page 19.
For details about the current system software update, see: “IMPORTANT - Install Latest OS
Updates, Patches, and Firmware” on page 14
IMPORTANT - Install Latest OS Updates, Patches,
and Firmware
Some product features are enabled only when the latest versions of operating systems, patches,
and firmware are installed. To retain optimal performance, security, and stability, you must
install the latest available operating systems, patches, and firmware.
System Software Release 1.11.0 is associated with system firmware version 3.2.10.21. Newer
versions have either a higher number, or a letter added. For example, a later system Software
Release might be associated with system firmware 3.2.11.0.
To verify that the server firmware version is a minimum of 3.2.10.21 or higher:
1.
Use Oracle ILOM to check your system firmware version.
■From the web interface, click System Information > Summary, then view the
System Firmware Version in the General Information table.
■From the CLI, type: show /System/Firmware or version.
14Sun Fire X4800 Server Product Notes • July 2017
Page 15
Diagnosing SAS Data Path Failures on Servers Using MegaRAID Disk Controllers
For more details, refer to information about viewing system information and inventory
in the Oracle ILOM Administrators Guide for Configuration and Maintenance, which is
available at http://www.oracle.com/goto/ilom/docs.
2.
Ensure that the server firmware version is at the minimum required version,
shown above, or a subsequent release, if available.
3.
If the required firmware (or newer) is not installed:
a.
Download the firmware from My Oracle Support at: https://support.oracle.
com
b.
Install the downloaded firmware.
Refer to the information about performing firmware updates in the Oracle ILOM
Administrators Guide for Configuration and Maintenance, which is available at http:
//www.oracle.com/goto/ilom/docs. Ensure that you perform the preparatory steps
described in that document before updating the firmware.
Note - Occasionally after installing the firmware, the Oracle ILOM web interface cannot
display the power state correctly on the power control page. To correct this problem, clear your
browser cache before logging in to the Oracle ILOM web interface.
Diagnosing SAS Data Path Failures on Servers Using
MegaRAID Disk Controllers
On Oracle x86 servers using MegaRAID disk controllers, Serial Attached SCSi (SAS) data path
errors can occur.
To triage and isolate a data path problem on the SAS disk controller, disk backplane (DBP),
SAS cable, SAS expander, or hard disk drive (HDD), gather and review the events in the disk
controller event log. Classify and analyze all failure events reported by the disk controller based
on the server SAS topology.
To classify a MegaRAID disk controller event, gather and parse the MegaRAID disk controller
event logs using the MegaCLI command:
For example, at the root prompt, type:
root# ./MegaCli64 adpeventlog getevents –f event.log aall
Success in AdpEventLog
Important Operating Notes15
Page 16
Oracle ILOM Deprecation Notice for IPMI 2.0 Management Service Present Behavior
Exit Code: 0x00
Note - Use the existing name of the event log as the name for the disk controller event log. This
produces a MegaRAID controller event log with the given file name event.log.
The following SCSI sense key errors found in the event log in SAS data path failures indicate a
SAS data path fault:
B/4B/05 :SERIOUS: DATA OFFSET ERROR
B/4B/03 :SERIOUS: ACK/NAK TIMEOUT
B/47/01 :SERIOUS: DATA PHASE CRC ERROR DETECTED
B/4B/00 :SERIOUS: DATA PHASE ERROR
A communication fault between the disk and the host bus adapter causes these errors. The
presence of these errors, even on a single disk, means there is a data path issue. The RAID
controller, SAS cables, SAS expander, or disk backplanes might be causing the interruption to
the communication in the path between the RAID controller and the disks.
Oracle Service personnel can find more information about the diagnosis and triage of hard
disk and SAS data path failures on x86 servers at the My Oracle Support web site: https://
support.oracle.com. Refer to the Knowledge Article Doc ID 2161195.1 If there are multiple,
simultaneous disk problems on an Exadata server, Oracle Service personnel can refer to
Knowledge Article Doc ID 1370640.1.
Oracle ILOM Deprecation Notice for IPMI 2.0 Management
Service Present Behavior
Present Behavior: IPMI 2.0 Sessions - Enabled (Default). Support for IPMI 2.0 client
interfaces.
Future Behavior: The following IPMI Management Service changes will occur in future
Oracle ILOM firmware releases after firmware version 3.2.7.
■
First feature change: Oracle ILOM will add a new client interface as an alternative to the
IPMI 2.0 Client interface.
■
Second feature change: The default configuration property for IPMI 2.0 Sessions will
change from Enabled to Disabled in a future release. Clients relying on IPMI 2.0 will be
unable to communicate with Oracle ILOM unless the configuration property for IPMI 2.0
Sessions is manually enabled.
16Sun Fire X4800 Server Product Notes • July 2017
Page 17
Oracle ILOM Deprecation Notice for Default Self-Signed Certificate
■
Third feature change: Removal of IPMI 2.0 client support. IPMI 2.0 clients will no longer
be able to communicate with Oracle ILOM.
For future updates about IPMI Management Service support in Oracle ILOM, refer to the latest
firmware release information in the Oracle ILOM Feature Updates and Release Notes forFirmware 3.2.x.
Oracle ILOM Deprecation Notice for Default Self-Signed
Certificate
Present Behavior: An earlier version of the default SSL self-signed certificate is provided by
Oracle ILOM.
Future Behavior: A newer version of the default SSL self-signed certificate will be provided in
a future Oracle ILOM firmware release.
Impact to Customer Configuration: After updating to a future firmware release, users
connecting to Oracle ILOM through the web interface will need to accept a newer version of the
default SSL self-signed certificate that is provided by Oracle ILOM. Customer provided SSL
certificates will not be impacted by this change.
For future updates about the default SSL self-signed certificate that is provided by Oracle
ILOM, refer to the latest firmware release information in the Oracle ILOM Feature Updatesand Release Notes for Firmware 3.2.x.
Broken Links in Sun Server X4800 Documentation Library
The following table lists the broken links in the Sun Server X4800 documentation library.
Note - The Sun Server X4800 Documentation Library includes a Sun Server X4800 HTML
Documentation Collection. This collection has the same chapters and content as the individual
documents in the following table. Many of the broken links listed below are also broken in the
corresponding sections of the Sun Server X4800 HTML Documentation Collection.
Red Hat Enterprise Linux -- https://access.redhat.com/certifications
■
SUSE Linux Enterprise Server -- https://www.suse.com/yessearch/Search.jsp
The following list shows the minimum supported operating systems.
■
Oracle Solaris 10 8/11, 10 9/10, 10 10/09 (64-bit)
Supported Software and Firmware19
Page 20
Supported Firmware
■
Oracle Solaris 11, 11.1, and 11.2 (64-bit)
■
Oracle Linux 5.5, 5.6, 5.7, 5.9, 5.10, 6.1, 6.4, 6.5 6.6, 7.0, and 7.1 (64-bit)
■
Oracle VM 2.2.1, 2.2.2, 3.0, 3.2.1, 3.3.1, 3.3.2
■
Red Hat Enterprise Linux 5.5, 5.6, 5.7, 6.0, 6.1, 6.0, 7.1
■
SUSE Enterprise Linux (SLES) 11 and SLES11 SP1 (64-bit)
■
Windows 2008 and R2 SP1 (64-bit)
■
VMware ESX/ESXi 4.0 U2, ESX/ESXi 4.1 U1, ESXi 5.0
Supported Firmware
SW 1.11.0 includes the following firmware:
■
Oracle ILOM -- 3.2.10.21 (r118444)
■
BIOS -- 11080300
SW 1.10.0 includes the following firmware:
■
Oracle ILOM -- 3.2.9.25 (r116305)
■
BIOS -- 11080300
Additional Software
Note - To obtain optimal performance, security, and stability, install system software release
1.11.0 or newer. Oracle recommends that you always install the latest available firmware.
The following additional software is available for download:
■
Oracle Hardware Installation Assistant (OHIA) 2.5
■
SunVTS 7.0 Patch Set 12 or later
Sun Java Enterprise System
Sun Java Enterprise System (Java ES) is a set of software components that provide services
needed to support enterprise-strength applications that are distributed across a network or
Internet environment. The Sun Java ES is preloaded on your server.
20Sun Fire X4800 Server Product Notes • July 2017
Page 21
Oracle Enterprise Manager Ops Center
Oracle Enterprise Manager Ops Center is a highly scalable, unified management platform for
physical and virtual environments. Use Oracle Enterprise Manager Ops Center to manage
multiplatform x86 and SPARC systems that are distributed throughout a global data center and
to integrate these systems with existing toolsets. Ready to facilitate many aspects of compliance
reporting (ITIL) and data center automation, Oracle Enterprise Manager Ops Center enables
management of thousands of systems simultaneously.
MegaRAID Storage Manager (MSM)
MegaRAID Storage Manager (MSM) is a configuration application that enables you to
configure, monitor, and maintain storage configurations on Integrated RAID controllers.
This graphical user interface (GUI) makes it easy for you to create and manage storage
configurations. There is also a MegaCLI command-line configuration utility, to manage your
RAID controllers. These applications are available on the Tools and Drivers image on the
product download site.
Oracle Enterprise Manager Ops Center
MSM enables you to easily configure the controllers, disk drives, and virtual disks on your
system. The Configuration wizard greatly simplifies the process of creating disk groups and
virtual disks. The Configuration wizard guides you through several simple steps to create your
storage configurations.
For more information on software for your integrated RAID controller, see: http://docs.
oracle.com/cd/E19418-01/index.html
SunVTS Bootable Diagnostics CD-ROM
The server is shipped with a bootable diagnostics CD-ROM. This CD-ROM is designed so that
the server boots using the Solaris OS on the CD-ROM and starts SunVTS software. Diagnostic
tests run and write output to log files that the service technician can use to determine the
problem with the server.
For information on SunVTS, see Using SunVTS Diagnostics Software in Sun Fire X4800Server Diagnostics Guide.
Note - SunVTS is now called Oracle VTS.
Supported Software and Firmware21
Page 22
Oracle Integrated Lights Out Manager (Oracle ILOM)
Oracle Integrated Lights Out Manager (Oracle ILOM)
Oracle ILOM is system management firmware that allows you to manage your server when the
host system is powered down. This is possible because Oracle ILOM runs on a separate Service
Processor (SP) that is powered by the host system’s standby power.
The following interfaces provide network access to the Oracle ILOM: command-line interface
(CLI), web interface, SNMP, and IPMI.
Oracle ILOM also supports remote access to the host’s system console through a network
remote keyboard video and mouse (RKVM). The host’s I/O to optical and floppy drives can be
redirected to real and virtual drives on the network. This allows a remote user to perform most
maintenance operations, including installing an operating system.
For more information on Oracle ILOM, refer to the Oracle ILOM 3.2 collection at: http://
www.oracle.com/goto/ilom/docs
Oracle Hardware Management Pack
Oracle Hardware Management Pack allows you to monitor hardware through the host operating
system either remotely using SNMP, or locally using command-line interface tools.
For more information, refer to the Oracle Hardware Management Pack Documentation Library
at http://www.oracle.com/goto/ohmp/docs.
Documentation Errata
This topic provides corrections to the Sun Fire X4800 documentation.
Use Oracle ILOM 3.2 Documentation After
Updating to System Software Release 1.7.0 or
Newer
When you update to system software release 1.7.0 or newer it updates Oracle ILOM as well.
Note the following documentation changes:
22Sun Fire X4800 Server Product Notes • July 2017
Page 23
Documentation Errata
Note - Oracle recommends you update to system software release 1.11.0 or newer as descibed
in “IMPORTANT - Install Latest OS Updates, Patches, and Firmware” on page 14.
■
The Oracle ILOM 3.0 collection no longer applies to your server. Instead refer to the Oracle
ILOM 3.2 collection at http://www.oracle.com/goto/ilom/docs.
■
The information in the Oracle Integrated Lights Out Manager (ILOM) 3.0 Supplement forthe Sun Fire X4800 is out of date. Instead, refer to the Oracle ILOM 3.2 collection.
Antistatic Wrist Straps Are Not Included With All
CRUs and FRUs
The service and installation documentation might state that antistatic wrist straps are included
with Customer Replaceable Units (CRUs) and Field Replaceable Units (FRUs). This is not
always true. Some CRUs and FRUs are shipped without antistatic wrist straps.
Previous Versions of the Service Manual
Incorrectly State That a FEM Is Only Supported on
CMOD 0 (CR 7107085)
In previous versions of the service manual, in the section Removing and Installing a Fabric
Expansion Module (CRU), the notes incorrectly state that the Fabric Expansion module (FEM)
is supported only in CMOD0 (BL 0).
These notes are incorrect. FEMs are supported on all CMODs.
The incorrect notes appear on pages 95 and 96. They have been fixed in the latest version of the
service manual.
Links to Oracle Websites Have Been Updated
Some printed documents have been reissued to update links to Oracle websites. The links in the
printed documents might be outdated, but they should be corrected in online documents. These
include:
■
Sun Fire X4800 Server Installation Guide
Supported Software and Firmware23
Page 24
Documentation Errata
■
Sun Fire X4800 Server Rack Mounting and Shipping Bracket User's Guide
Links to documentation have been moved to: http://docs.oracle.com/cd/E19140-01/index.
html
Name Correction for Oracle Integrated Lights Out
Manager
The Oracle Integrated Lights out Manager (Oracle ILOM) might be referred to as the Sun
Integrated Lights Out Manager in some documents and screen displays.
Front Mounting Bracket Illustration from
Installation Guide is Incorrect
A figure in How to Install the Rack Mounting Hardware in a Round-Hole Rack in Sun Fire
X4800 Server Installation Guide is incorrect. It shows the screws for installing the front
mounting bracket in a threaded rack incorrectly being inserted from the front of the rack.
Instead, the screws should be inserted from the back of the rack.
The following figure shows the front mounting brackets for both square-hole racks (1) and for
threaded racks (2). In the Installation Guide and in this figure, (1) is correct and unchanged.
However, in the Installation Guide, (2) is incorrect. The following figure shows (2) correctly.
24Sun Fire X4800 Server Product Notes • July 2017
Page 25
■
This figure is incorrect in 821–0285–10, the printed version of this document.
■
They are shown correctly in 821–0285–11, the online version of this document.
Documentation Errata
Supported Software and Firmware25
Page 26
26Sun Fire X4800 Server Product Notes • July 2017
Page 27
Oracle x86 Products Accessibility
This section describes the accessibility features that are part of Oracle x86 hardware, firmware,
and related documentation.
Oracle strives to make its products, services, and supporting documentation usable and
accessible to the disabled community. To that end, products, services, and documentation
include features that make the product accessible to users of assistive technology.
For more information about Oracle's commitment to accessibility, go to:
Oracle x86 hardware has color-coded labels, component touch points, and status indicators
(LEDs) that provide information about the system. These labels, touch points, and indicators
can be inaccessible features for sight-impaired users. The product HTML documentation
provides context and descriptive text available to assistive technologies to aid in interpreting
status and understanding the system. System-level descriptions and status indicator
interpretation can be found in the product Service Manual. The documentation also provides
diagrams and screenshots that do not rely on color. Within the diagrams, callouts indicate the
referenced component information. The callout descriptions are mapped within a table. All
images and tables in the documentation include descriptive alternative text.
Another method to obtain information about the system is to use the built-in Oracle Integrated
Lights Out Manager (ILOM). Oracle ILOM provides a browser-based interface and a
command-line interface that support assistive technologies for real-time viewing of system
status, indicator interpretation, and system configuration. For details, see "Oracle ILOM
Accessibility."
You can access the accessible HTML documentation for Oracle x86 hardware products at:
■
http://docs.oracle.com/en/servers/
Oracle x86 Products Accessibility27
Page 28
Oracle ILOM Accessibility
Oracle ILOM Accessibility
You can use the Oracle Integrated Lights Out Manager (ILOM) browser user interface
(BUI) to monitor and manage the server hardware. The Oracle ILOM BUI does not require
a special accessibility mode; rather, its accessibility features are always available. The BUI
was developed using standard HTML and JavaScript and its features conform to accessibility
guidelines.
To navigate a BUI page and select items or enter commands, you can use standard keyboard
inputs, such as using the Tab key to go to a selection, or the up and down arrow keys to
scroll through the page. You can also make menu selections by using standard keyboard
combinations.
For example, using the Oracle ILOM Open Problems BUI page, you can identify faulted
memory modules (DIMMs) or processors (CPUs) that would otherwise be identified by a lit
LED indicator on the motherboard. Likewise, you can use the Oracle ILOM BUI to monitor the
hardware power states that are also indicated by flashing LED indicators on the hardware.
The Oracle ILOM command-line interface (CLI) is an alternative and equivalent way to access
the Oracle ILOM BUI features and functionality. Because the operating systems that run on
the Oracle server hardware support assistive technologies to read the content of the screen,
you can use the CLI as an equivalent means to access the color-based, mouse-based, and other
visual-based utilities that are part of the BUI. For example, you can use a keyboard to enter CLI
commands to identify faulted hardware components, check system status, and monitor system
health.
You can use the Oracle ILOM Remote Console Plus to access both a text-based serial console
and a graphics-based video console that enable you to remotely redirect host server system
keyboard, video, mouse, and storage devices. Note, however, that the Oracle ILOM Java
Remote Console does not support scaling of the video frame within the Java application. You
need to use assistive technology to enlarge or reduce the content in the Java Remote Console
Plus display.
As an alternative method to using the BIOS Setup Utility to configure BIOS settings,
Oracle ILOM provides a set of configurable properties that can help you manage the BIOS
configuration parameters on an Oracle x86 server. Using Oracle ILOM, you can:
■
Back up a copy of the BIOS configuration parameters to an XML file using the Oracle
ILOM BUI.
■
Edit the XML file using a standard XML editor. The BIOS XML tags correlate directly to
the BIOS screen labels.
■
Restore the XML file of the backed up or edited configuration parameters to BIOS.
The BUI and CLI methods for using Oracle ILOM are described in the accessible HTML
documentation for Oracel ILOM at:
28Sun Fire X4800 Server Product Notes • July 2017
Page 29
Oracle Hardware Management Pack Accessibility
■
http://www.oracle.com/goto/ilom/docs
Oracle Hardware Management Pack Accessibility
Oracle Hardware Management Pack software is a set of command-line interface (CLI) tools.
Oracle Hardware Management Pack software does not include product-specific accessibility
features. Using a keyboard, you can run the CLI tools as text commands from the operating
system of a supported Oracle server. All output is text-based.
Additionally, most Oracle Hardware Management Pack tools support command output to a text
log file or XML file, which can be used for text-to-speech conversion. Accessible manual pages
(man pages) are available that describe the Hardware Management Pack tools on the system on
which those tools are installed.
Installation and uninstallation of Oracle Hardware Management Pack can be performed
manually, using text commands entered from the CLI. Assistive technology products such as
screen readers, digital speech synthesizers, or magnifiers can be used to read the content of the
screen.
Refer to the assistive technology product documentation for information about operating system
and command-line interface support.
The CLI tools for using the software are described in the accessible HTML documentation for
Hardware Management Pack at:
■
http://www.oracle.com/goto/ohmp/docs
BIOS Accessibility
When viewing BIOS output from a terminal using the serial console redirection feature, some
terminals do not support function key input. However, BIOS supports the mapping of function
keys to Control key sequences when serial redirection is enabled. Descriptions of the function
key to Control key sequence mappings are provided in the product documentation, typically
within the server Service Manual. You can navigate the BIOS Setup Utility by using either a
mouse or keyboard commands.
As an alternative method of configuring BIOS settings using the BIOS Setup Utility screens,
Oracle ILOM provides a set of configurable properties that can help you manage the BIOS
configuration parameters on an Oracle x86 server. For more information, see "Oracle ILOM
Accessibility."
Oracle x86 Products Accessibility29
Page 30
Documentation Accessibility
BIOS information and its functions are typically documented in the product Service Manual or
Installation Guide.
Documentation Accessibility
Documentation for Oracle hardware is provided in HTML and PDF formats. The HTML
documents are accessible using standard operating system controls and assistive technology.
PDF documents are also provided; however, PDF is not an accessible format. PDF documents
are considered support documents because the PDF content is available in accessible HTML
format.
Product documentation provides figures, other types of images, and screenshots that do not
rely on color for interpretation. Within the figures, callouts indicate the referenced component
information. The callouts are mapped within a table to provide text descriptions of the
referenced parts of the figures. In addition, alternative text is provided for all tables and images
that provides the context of the information and images.
Note that screen readers might not always correctly read the code examples in the
documentation. The conventions for writing code require that closing braces should appear on
an otherwise empty line. However, some screen readers might not always read a line of text that
consists solely of a bracket or brace.
The documentation might contain links to web sites of other companies and organizations
that Oracle does not own or control. Oracle neither evaluates nor makes any representations
regarding the accessibility of these web sites.
You can access the accessible HTML documentation for Oracle x86 products at:
■
http://docs.oracle.com/en/servers/
30Sun Fire X4800 Server Product Notes • July 2017
Page 31
Hardware Issues
The following table lists the current hardware issues.
Hardware IssuesWorkaround
“System Posts CPU Fault and Fails to Boot On System With Six Sun InfiniBand Dual Port
4x QDR PCIe Low Profile Host Channel Adapter M2 Cards (22536804)” on page 32
“Reset Takes a Long Time and Causes the Server to Power Cycle” on page 32N/A
“Update SP Firmware After Replacing a CMOD or SP Module (15465497)” on page 33Yes
“MegaCLI CfgEachDskRaid0 Command Does Not Work Correctly (CR
7121867)” on page 33
“DIMM Mismatch on Odd Numbered Pairs Results in DDR Training Failed on Entire
Branch (CR7101624 and CR7111545)” on page 34
“If DIMM Failure Causes Branch Failure, Oracle ILOM Might Not Identify Them as Faulty
(CR7099038 and CR7111543)” on page 35
“When Inserting a CMOD, Simultaneously Rotate Handles Until They Touch the
Chassis” on page 35
“Video Output of CMM Is Distorted or Missing” on page 35Yes
“SATA Drives Not Accessible Through NEM1 (CR 7003993)” on page 36Yes
“Supported Rack Mounting Configuration” on page 36N/A
“DIMM Population Rules” on page 36N/A
“Supported Rack Mounting Configuration” on page 36N/A
“Locking Screws Secure Shelf Rails to Rear Mounting Brackets” on page 39Yes
“Message About Missing Disks After Failed Boot (CR 6939507)” on page 41Yes
“System Does Not Power Up After Power Cycling (CR 6950414)” on page 41Yes
“System Does Not Recognize DIMM Due to SPD Corruption (CR
6946921)” on page 41
Yes
Yes
Yes
Yes
Yes
Yes
Hardware Issues31
Page 32
System Posts CPU Fault and Fails to Boot On System With Six Sun InfiniBand Dual Port 4x QDR PCIe Low Profile Host Channel
Adapter M2 Cards (22536804)
System Posts CPU Fault and Fails to Boot On System With
Six Sun InfiniBand Dual Port 4x QDR PCIe Low Profile Host
Channel Adapter M2 Cards (22536804)
A system equipped with six Sun InfiniBand Dual Port 4x QDR PCIe Low Profile Host Channel
Adapter M2 Cards might post CPU faults and fail to boot.
Note - The Sun Fire X4800 supports a maximum of six Sun InfiniBand Dual Port 4x QDR PCIe
Low Profile Host Channel Adapter M2 Cards.
Workaround
1. Access the BIOS Setup Utility.
2. Navigate to RC Settings > QPI and change:
MMIOH Size per IOH from 2Gb to 4Gb (the default is 2Gb).
3. Navigate to RC Settings > Chipset > NorthBridge Configuration and change:
PCIE-MMIO-64 Bits Support to Enabled (the default is Disabled).
4. Save your changes and exit the BIOS Setup Utility.
5. Reboot the server.
Reset Takes a Long Time and Causes the Server to Power
Cycle
If you have a pending BIOS upgrade, a routine reset might take longer than expected and might
cause your server to power cycle and reboot several times. This is expected behavior as it is
necessary to power cycle the server to upgrade the BIOS firmware. If the upgrade includes an
FPGA update, it can take more than 30 minutes to complete.
A pending BIOS upgrade exists when both conditions are true:
■
You update the BIOS and service processor firmware using Oracle ILOM.
■
You select the option to Delay BIOS Upgrade.
■
The host is powered on.
If you then reboot the server expecting a routine server reset and instead initiate a (delayed)
BIOS upgrade, wait until the upgrade is finished. Do not interrupt the process, as this can result
in corrupted firmware and server down time.
32Sun Fire X4800 Server Product Notes • July 2017
Page 33
Update SP Firmware After Replacing a CMOD or SP Module (15465497)
On older versions of Oracle ILOM, a pending BIOS upgrade does not take place until you
power cycle your server.
On newer versions of Oracle ILOM, a pending BIOS upgrade takes place when you reset your
system.
Caution - If you reboot the server expecting a routine server reset and instead initiate a
(delayed) BIOS upgrade, wait until the upgrade is finished. Do not interrupt the process as this
can result in corrupted firmware and server down time.
Note - Oracle ILOM and BIOS updates are designed to work together. When you have a
pending BIOS upgrade, it is recommended that you install the upgrade by resetting or power
cycling your server as soon as possible.
Update SP Firmware After Replacing a CMOD or SP Module
(15465497)
When the replacement of either a CMOD or the service processor results in an incompatibilty
between the hardware revision of the component and the firmware version of either the SP or
the BIOS, the recommendation is to maintain compatibility with the SP firmware. Therefore,
update or downgrade the system firmware package to the version compatible with the SP.
MegaCLI CfgEachDskRaid0 Command Does Not Work
Correctly (CR 7121867)
The CfgEachDskRaid0 command has two problems when run in the preboot CLI:
■
It assigns newly created logical volumes in apparently random order.
■
It sometimes fails to create all volumes.
These problems only occur when the command is run in the preboot environment. Normally,
you only configure a single boot volume in the preboot CLI anyway.
Workaround
This workaround allows you to configure the drives individually using the CfgLdAdd command
in the preboot CLI.
Hardware Issues33
Page 34
DIMM Mismatch on Odd Numbered Pairs Results in DDR Training Failed on Entire Branch (CR7101624 and CR7111545)
1. Run the following command to determine your enclosure number:
-pdlist -a0 | more
Adapter #0
Enclosure Device ID: 252
Slot Number: 0
Note - The typical enclosure number is 252, unless there are more than eight drive slots.
2. Use the CfgLdAdd command to build RAID volumes for each disk. For example:
-CfgLdAdd -r0[252:0] -a0
-CfgLdAdd -r0[252:1] -a0
-CfgLdAdd -r0[252:2] -a0
-CfgLdAdd -r0[252:3] -a0
Note - If your enclosure number is not 252, substitute your enclosure number in these
commands.
DIMM Mismatch on Odd Numbered Pairs Results in
DDR Training Failed on Entire Branch (CR7101624 and
CR7111545)
If a DIMM mismatch error (size, rank, or speed) occurs on odd-numbered DIMM pairs, it might
cause a ddr-training-failed error on the entire branch, thus disabling all DIMMs on that
branch.
Furthermore, Oracle ILOM might not mark them as faulty or send a diagnostic message, but
only send one ereport on boot:
ereport.memory.mrc.init-no-memory@/sys/blX/pX/mrX
Workaround
Use the Oracle ILOM Fault Manager to view information about active faults (fmadm faulty) to
identify the mismatched DIMM pair.
34Sun Fire X4800 Server Product Notes • July 2017
Page 35
If DIMM Failure Causes Branch Failure, Oracle ILOM Might Not Identify Them as Faulty (CR7099038 and CR7111543)
Then replace the pair with a matched pair. If the fault isn't cleared automatically, you must
manually clear it using the appropriate Oracle ILOM fmadm repair command as described in the
Oracle ILOM documentation.
If DIMM Failure Causes Branch Failure, Oracle ILOM Might
Not Identify Them as Faulty (CR7099038 and CR7111543)
When fault occurs on DIMMs that result in all DIMMs on the same branch being disabled,
Oracle ILOM might not mark them as faulty or send diagnostic message, but only send one
ereport:
ereport.memory.mrc.init-no-memory@/sys/blX/pY/mrZ
Where X is the blade slot number, Y is the processor, and Z is the DIMM memory branch.
Workaround
Use the Oracle ILOM Fault Manager to view information about active faults (fmadm faulty) to
identify the mismatched DIMM pair.
Then replace the pair with a matched pair. If the fault isn't cleared automatically, you must
manually clear it using the appropriate Oracle ILOM fmadm repair command as described in the
Oracle ILOM documentation.
When Inserting a CMOD, Simultaneously Rotate Handles
Until They Touch the Chassis
When you insert a CMOD into the chassis, simultaneously rotate both handles in past the locks
until they touch the chassis before releasing them. It is important that the handles actually touch
the chassis to ensure that the pawl completes the action of seating the CMOD connectors.
Video Output of CMM Is Distorted or Missing
Under certain rare conditions, the video output from the CMM might be distorted, colored
incorrectly, or not visible. This occurs when a video monitor is connected to the video port on
the multiport cable.
Hardware Issues35
Page 36
SATA Drives Not Accessible Through NEM1 (CR 7003993)
Workaround
This condition is benign, and if the video display is acceptable, you can ignore it.
1. Connect to the video display using Oracle ILOM.
For instructions for using Oracle ILOM to view the video display, refer to CommunicatingWith the ILOM and the System Console in Sun Fire X4800 Server Installation Guide.
SATA Drives Not Accessible Through NEM1 (CR 7003993)
SATA drives only work with NEM0. If NEM0 is not present, the system is unable to see or
access SATA drives.
This is expected behavior. Multipathing is not supported for SATA drives.
Workaround
If your system has SATA drives, make sure that NEM0 is present.
Supported Rack Mounting Configuration
You can install up to eight Sun Fire X4800 servers in a Sun Rack II 1042/1214 as long as your
datacenter meets the cooling requirements to support these systems.
DIMM Population Rules
This section provides the rules for adding and replacing DIMMs on a CMOD. It updates and
enhances the information in DIMM Population Rules in Sun Fire X4800 Server Service Manual.
For additional information, including how to prepare the system for servicing, how to remove
a CMOD, and how to remove or replace DIMMs, see Removing and Installing Components in
Sun Fire X4800 Server Service Manual.
Each CMOD has 32 DIMM slots, with 16 allocated to each CPU. CMODs are shipped with
configurations of 16 or 32 DIMMs, meaning that each CPU has either 8 or 16 DIMMs.
■
On a CMOD with 16 DIMMs, you can replace DIMMs, and you can add additional
DIMMs.
■
On a CMOD with 32 DIMMs, you can replace DIMMs, but because they are fully
populated, you cannot add DIMMs.
36Sun Fire X4800 Server Product Notes • July 2017
Page 37
DIMM Population Rules
The rules for adding and replacing DIMMs are:
■
When you replace DIMMs, you must match parameters.
■
When you add DIMMs, you must match parameters and populate them in the specified
order.
Matching Parameters
There are four parameters to consider when replacing DIMMs: size, speed, density, and
manufacturer's model number.
■
All of the DIMMs in the CMOD must be the same size, speed, and density.
■
Both of the DIMMs in a pair must be the same size, speed, and density, and have the same
manufacturer's model number. The pairs are 0/4, 1/5, 2/6, 3/7, 8/12, 9/13, 10/14, and 11/15.
The following table shows the requirements.
Parameter and ValueMatch Within CMODMatch Within Pairs
Size — 2 GB, 4 GB, or 8 GBxx
Speed — 1066 MHz JEDEC or DDR3 ECC RDIMMsxx
Density — Single or dual rankxx
Manufacturer's model numberx
DIMM Population Order
You can add DIMMs according to the following rules:
■
You cannot add DIMMs to a 32–DIMM system because it is already fully populated.
■
All DIMMs must have their parameters matched, as described above.
■
Only four or eight DIMM upgrades are supported. As a result, all color blocks must be
either full or empty.
■
DIMMs must be populated in the correct order.
On a 16–DIMM CMOD, all the blue and white slots should already be populated.
When you add DIMMs, fill the black slots (D1/D5 and D10/14) first, and fill the green slots
(D3/D7 and D11/D15) last.
OrderNumbersSlot ColorMinimum Configuration
1D0/D4Blue
These should be filled by default
Hardware Issues37
Page 38
DIMM Population Rules
The following figure shows the location of the DIMMs on the CMOD.
OrderNumbersSlot ColorMinimum Configuration
2D8/D12Blue
3D2/D6White
4D10/D14White
5D1/D5Black
6D9/D13Black
7D3/D7Green
8D11/D15Green
Fill these first
Fill these last
38Sun Fire X4800 Server Product Notes • July 2017
Page 39
Locking Screws Secure Shelf Rails to Rear Mounting Brackets
LegendDescription
1CPU 0
2CPU 1
3Towards the front of the CMOD
Locking Screws Secure Shelf Rails to Rear Mounting
Brackets
A pair of locking screws (provided) secure the shelf rails to the rear mounting brackets. These
screws prevent the mounting brackets from being accidentally disengaged, which could cause
the server to fall out of the rack.
How to Install the Locking Screws
Before You Begin
1.
Install the server in the rack, as described in Installing the Server in a Rack Using the Universal
Rack Mounting Kit in Sun Fire X4800 Server Installation Guide.
From the rear of the server, insert the screw (1) through the rack (2) and the rear
mounting bracket (3) so that it protrudes from the mounting bracket just above
the flange on the shelf rail (5).
■
For a round hole (tapped) rack, use either an M6 x 12 or a 10–32 x 7/16 screw.
■
For a square hole (untapped) rack (shown below), use an M6 cage nut (4) and an M6 x 16
screw.
The following figure shows the details for a square-hole (untapped) rack.
Hardware Issues39
Page 40
How to Install the Locking Screws
Note - The shelf rails must be inserted into the mounting bracket before performing this
procedure. They are shown separated in this picture to highlight the alignment.
Figure Legend
1Screw
2Cage nut (used for square hole racks only)
3Rack
4Rear mounting bracket
5Shelf rail
2.
Repeat Step 1 for the other side.
40Sun Fire X4800 Server Product Notes • July 2017
Page 41
Message About Missing Disks After Failed Boot (CR 6939507)
Message About Missing Disks After Failed Boot (CR
6939507)
On rare occasions, after initial power up, the system fails to boot and displays a message:
All of the disks from your previous configuration are gone.
If this is an unexpected message, then please power off your system and check
your cables to ensure all disks are present. Press any key to continue, or "C"
to load the configuration utility.
Workaround
1. Power cycle the system.
The system should boot normally.
2. If the system still does not boot, and the same message appears, contact Oracle customer
service.
System Does Not Power Up After Power Cycling (CR
6950414)
Under rare conditions, when the system is power-cycled, it might not power up.
Workaround
Remove and reapply AC power to and from the system.
Either:
■
Switch the AC power Off then switch it back On.
■
Remove all the AC power cords from the power supplies and then plug them back in.
■
Remove the power supplies and then replace them.
System Does Not Recognize DIMM Due to SPD Corruption
(CR 6946921)
Under rare conditions, a DIMM can get corrupted. When this happens, the system no longer
recognizes that DIMM, and as a result, the system has less memory available to it than what
was configured.
Hardware Issues41
Page 42
How to Identify a Corrupted DIMM
How to Identify a Corrupted DIMM
1.
Log on to Oracle ILOM.
2.
Enter the command to display faulted DIMMs.
■From the web interface, select the Fault Management tab.
■
From the CLI, enter show/SP/faultmgt.
Oracle ILOM displays the faulted DIMMs with the fault class. It might show one of the
following:
If the fault class is anything else besides dimm_ce or dimm_ue, contact your Oracle
service representative.
42Sun Fire X4800 Server Product Notes • July 2017
Page 43
Oracle Solaris 10 Operating System Issues
The following table lists the Oracle Solaris OS issues.
Current Solaris OS IssuesWorkaround
“Oracle Solaris Installation Takes a Long Time On Systems With Four or More
Sun InfiniBand Dual Port 4x QDR PCIe Low Profile Host Channel Adapter M2
Cards” on page 43
“System Panics During Solaris 10 10/09 OS Installation When Solaris Attempts to Boot (CR
6880963)” on page 45
“System Panics or Hangs During Solaris OS Installation (CR 6920274)” on page 46Yes
“Messages Complain of Failure to Allocate I/O Resources (CR 6984329)” on page 47Yes
“Patch Required for Solaris FMA” on page 47Yes
“Solaris MSI Interrupts Are Depleted (CR 6669984 )” on page 48Yes
“System Might Panic With "unowned mutex" Message (CR 6893274)” on page 49No
“Hotplugging PCIe Express Modules in Slots 2.0 or 2.1 Might Not Work (CR
6954869)” on page 49
“A System With a Combo GbE/8Gb FC Express Module HBA Might Get a BAD TRAP
Panic (CR 6942158)” on page 49
Yes
Yes
Yes
Yes
Oracle Solaris Installation Takes a Long Time On Systems
With Four or More Sun InfiniBand Dual Port 4x QDR PCIe
Low Profile Host Channel Adapter M2 Cards
If you install Oracle Solaris on a system with four or more Sun InfiniBand Dual Port 4x QDR
PCIe Low Profile Host Channel Adapter M2 Cards, the installation can take much longer than
normal.
Note - The Sun Fire X4800 server supports a maximum of six Sun InfiniBand Dual Port 4x
1. Remove the Sun InfiniBand Dual Port 4x QDR PCIe Low Profile Host Channel Adapter
M2 Cards.
2. Complete the Oracle Solaris installation.
3. Reinstall the Sun InfiniBand Dual Port 4x QDR PCIe Low Profile Host Channel Adapter
M2 Cards
Solaris 10 Operating System Overview
The Solaris 10 Operating System (OS) delivers the security, manageability, and performance
that IT professionals need to help increase service levels and decrease costs and risk. It also
serves as the foundation for the Sun Java Enterprise System, an integrated, open, standardsbased software system delivered using a new predictable approach for development, testing, and
servicing.
If the Solaris OS is preinstalled on your server, it already includes required patches. If you use
your own custom OS install environment, you must update your install environment with the
following patches, which can be downloaded from https://support.oracle.com.
■
Solaris 10 9/10 requires patch 146025–01 or later. It is recommended that you apply the
latest patch cluster.
■
Solaris 10 10/09 is supported for existing installations but deprecated for new installations.
Note - It is highly recommended that you upgrade to Solaris 10 9/10 or install the latest
Solaris 10 patch cluster.
Solaris 10 10/09 requires the following patches:
■
142901–09 or higher. This addresses the following issues:
■
“System Panics or Hangs During Solaris OS Installation (CR
6920274)” on page 46
■
“Patch Required for Solaris FMA” on page 47
■
SunVTS 7.0ps08 or later
■
143958–03 or later if you have a ComboGbE/8Gb FC Express Module
■
144374–01 or later
■
Other patches might be available as well.
44Sun Fire X4800 Server Product Notes • July 2017
Page 45
System Panics During Solaris 10 10/09 OS Installation When Solaris Attempts to Boot (CR 6880963)
System Panics During Solaris 10 10/09 OS Installation
When Solaris Attempts to Boot (CR 6880963)
When installing Solaris 10 10/09, the system panics if it has a large memory size
(approximately 768 GB or larger)
For example:
panic[cpu0]/thread=fffffffffbc288a0: mutex_enter: bad mutex,
lp=18 owner=f00006f7f0000748 thread=fffffffffbc288a0
Note - If you are doing a PXE install, you might choose to add patch 142091–09 or higher
to your install image. This allows you to complete the installation without reducing the
memory size. However, you must still reduce the memory size (step 2) and add the patch to
the installed system (step 4) before you can boot it.
2. Reduce the system memory to 512 GB.
We suggest you remove CMODs 3 and 4. You can use other techniques if desired.
For details, refer to Adding, Removing, and Installing a CMOD (CRU) in Sun Fire X4800Server Service Manual.
3. Install the operating system.
This is not necessary if your system already has an operating system installed.
4. Install patch 142901–09 or higher.
5. Replace the memory.
If you removed the CMODs, for details, refer to Adding, Removing, and Installing a CMOD(CRU) in Sun Fire X4800 Server Service Manual.
Oracle Solaris 10 Operating System Issues45
Page 46
System Panics or Hangs During Solaris OS Installation (CR 6920274)
System Panics or Hangs During Solaris OS Installation (CR
6920274)
If you try to do a network install using a 32–bit version of the Solaris OS on the Sun Fire
X4800 server, it might fail. It might panic or hang during installation.
32–bit versions of the Solaris OS are not supported. When you install the Solaris OS from a
PXE server, the netboot environment must be configured to use the 64–bit kernel.
When a 32–bit install environment is used, it displays a message similar to the following:
SunOS Release 5.10 Version Generic 141445-09 32-bit
Copyright 1983-2009 Sun Microsystems, Inc. All rights reserved
Workaround
You can fix this by changing the PXE boot menu to include amd64 on the kernel and module
lines.
The following displays show examples of the lines before and after this change. The lines have
been wrapped to make the display fit the page.
Messages Complain of Failure to Allocate I/O Resources (CR 6984329)
Messages Complain of Failure to Allocate I/O Resources
(CR 6984329)
Messages that complain about failures to allocate I/O resources might appear in POST and in
log files.
For example, you might see:
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.0
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.1
Workaround
BIOS tries to allocate I/O resources. If it is unable to allocate I/O resources successfully, it
displays error messages.
The OS might try as well. However if it tries, it fails and generates additional failure messages
because it uses the same method as BIOS. However, most drivers can manage this condition.
1. Usually, you can ignore these messages.
2. If you continue to experience I/O resource issues, see I/O and Interrupt Resource Allocation
in the Sun Fire X4800 Server Installation Guide.
Patch Required for Solaris FMA
For Solaris 10 10/09, in order to use Solaris Fault Management Architecture (FMA) with your
server, you must install patch 142901–09 after installing the OS.
This patch is also recommended to utilize the maximum number of interrupts on the server.
Workaround 1
Upgrade to Solaris 10 9/10 or newer.
Workaround 2
1. Check to see if you have installed patch 142901–09 or later. If not:
a. Download patch 142901–09 or later from:
https://support.oracle.com
b. Follow the patch instructions to install the patch.
2. Issue the following commands:
Oracle Solaris 10 Operating System Issues47
Page 48
Solaris MSI Interrupts Are Depleted (CR 6669984 )
rem_drv intel_nhmex
add_drv -i '"pci8086,3438"' intel_nhmex
3. Reboot your server.
Solaris MSI Interrupts Are Depleted (CR 6669984 )
The Solaris OS might run out of MSI interrupts on 8-socket systems.
The Solaris OS is designed to allocate 32 interrupt vectors at priority level six. Also, each
network device is assigned two interrupts. If the system contains more than 16 network devices,
it will exhaust the supply of interrupts on level six, leaving some devices non-functional.
When this occurs, error messages similar to the following appears:
■
On the console:
Feb 25 15:45:04 mpk12-3214-189-156 pcplusmp: WARNING: No interrupt vector:
pciex8086,10f7 instance 1
Feb 25 15:45:04 mpk12-3214-189-156 pcplusmp: WARNING: Sharing vectors:
pciex8086,10f7 instance 1 and SCI
Feb 25 15:45:06 mpk12-3214-189-156 pcplusmp: WARNING: No interrupt vector:
pciex8086,10f7 instance 5
Feb 25 15:45:06 mpk12-3214-189-156 pcplusmp: WARNING: Sharing vectors:
pciex8086,10f7 instance 1 and pciex8086,10f7 instance 5
■
In /var/adm/messages:
Feb 25 15:44:53 mpk12-3214-189-156 ixgbe: [ID 611667 kern.info]
Feb 25 15:44:53 mpk12-3214-189-156 ixgbe: [ID 611667 kern.info]
NOTICE: ixgbe7: MSI-X not used, force rx and tx queue number to 1
Note - If you have not installed patch 142901–09 or later, the Solaris OS is currently limited to
allocating 31 vectors at priority level six, meaning that it can only support 15 network devices
on level six.
Workaround 1
For Solaris 10 9/10, install patch 146025–03 or higher.
This fixes the problem in most cases.
48Sun Fire X4800 Server Product Notes • July 2017
Page 49
System Might Panic With "unowned mutex" Message (CR 6893274)
Note - This patch requires Solaris 10 9/10 and cannot be used with Solaris 10 10/09.
Workaround 2
When you encounter a shortage of I/O interrupts on level 6, Oracle recommends that you assign
one of the drivers to interrupt level 5 or 4.
■
Interrupt level 5 is the first priority
■
Interrupt level 4 is next
See Allocating MSI Interrupt Space (Oracle Solaris OS Only) in the Sun Fire X4800 ServerInstallation Guide for corrective actions.
System Might Panic With "unowned mutex" Message (CR
6893274)
Under rare conditions, the system might panic and display the message:
turnstile_block: unowned mutex
This is a known Solaris OS issue.
Hotplugging PCIe Express Modules in Slots 2.0 or 2.1 Might
Not Work (CR 6954869)
On an 8–socket system, hotplugging PCIe express modules in slots 2.0 or 2.1 might not work.
This is due to a possible shortage of hotplug interrupts on the system.
Workaround
Insert these modules before booting the system.
A System With a Combo GbE/8Gb FC Express Module HBA
Might Get a BAD TRAP Panic (CR 6942158)
Due to a known issue in the Solaris OS, a system with a Combo GbE/8Gb FC Express Module
HBA might panic during boot with a BAD TRAP message.
Oracle Solaris 10 Operating System Issues49
Page 50
A System With a Combo GbE/8Gb FC Express Module HBA Might Get a BAD TRAP Panic (CR 6942158)
This issue applies to Solaris 10 10/09. It is fixed in Solaris 10 9/10.
Workaround
Install patch number 143858–03 or newer, or install Solaris 10 9/10.
50Sun Fire X4800 Server Product Notes • July 2017
Page 51
Linux Issues
The following table lists the issues with related to the Linux operating system on the Sun Fire
X4800 Server server.
Linux IssuesWorkaround
“Oracle and Red Hat Linux 7.2 Installation Fails With AER Error on Systems With Four or
More Sun InfiniBand Dual Port 4x QDR PCIe Low Profile Host Channel Adapter M2 Cards
(22686146)” on page 52
“False MCE Errors Appear in /var/log/mcelog (CR 7104293)” on page 52
“I/O Does Not Work On SLES11 SP1 with XEN (CR 6965290 and
CR7110443)” on page 53
“RHEL6 Kdump Runs Out of Memory (CR 7000993 and CR 7000942)” on page 53Yes
“SLES11 SP1 Does Not Produce A Kdump File After A Crash (CR7001706)” on page 54Yes
“SLES11 Fails to Install On Systems Equipped with 6–Core CMODs (CR
7024769)” on page 54
“Messages Complain of Failure to Allocate I/O Resources (CR 6984329)” on page 55Yes
“Xen Profiler Is Not Supported in Oracle Linux 5.5 (CR 6839366)” on page 55No
“Oracle Linux 5.5 Does Not Support CPU Throttling (CR 6847286)” on page 55No
“InfiniBand PCIe Module Does Not Function Correctly With SLES 11 (CR 6909512, CR
6921163)” on page 56
“LEDs on PCIe ExpressModule Work Incorrectly with Oracle Linux 5.5 (CR
6894954)” on page 56
“Oracle Linux 5.5 Xen Kernel Hangs Due to Long Timeout (CR 6899040)” on page 57Yes
“10GbE PCIe Card in 8-Socket System with SLES11 Can Cause Kernel to Crash (CR
6912821)” on page 58
Yes
N/A
Yes
Yes
Yes — Fixed in SP1
No
Yes
Linux Issues51
Page 52
Oracle and Red Hat Linux 7.2 Installation Fails With AER Error on Systems With Four or More Sun InfiniBand Dual Port 4x QDR
PCIe Low Profile Host Channel Adapter M2 Cards (22686146)
Oracle and Red Hat Linux 7.2 Installation Fails With AER
Error on Systems With Four or More Sun InfiniBand Dual
Port 4x QDR PCIe Low Profile Host Channel Adapter M2
Cards (22686146)
Oracle Linux 7.2 or Red Hat Linux 7.2 installation might fail with an AER error on a system
with four or more Sun InfiniBand Dual Port 4x QDR PCIe Low Profile Host Channel Adapter
M2 Cards.
Note - The Sun Fire X4800 server supports a maximum of six Sun InfiniBand Dual Port 4x
MMIOH Size per IOH from 2Gb to 4Gb (the default is 2Gb).
3. Navigate to RC Settings > Chipset > NorthBridge Configuration and change:
PCIE-MMIO-64 Bits Support to Enabled (the default is Disabled).
4. Save your changes and exit the BIOS Setup Utility.
5. Finish the operating system installation.
False MCE Errors Appear in /var/log/mcelog (CR 7104293)
Occasionally, many errors appear in the MCE log, /var/log/mcelog. For example:
CPU 0 BANK 9 MCG status:
MCi status:
Error enabled
MCA: MEMORY CONTROLLER GEN_CHANNELunspecified_ERR
Transaction: Generic undefined request
STATUS 900000400009008f MCGSTATUS 0
■
If this is a one-time or an occasional event, you can ignore it. It should not impact
performance.
■
If the errors persist, contact Oracle Services for further review. Open a support case with
Oracle Services and reference (Doc ID #1459158.1) for further analysis and evaluation.
52Sun Fire X4800 Server Product Notes • July 2017
Page 53
I/O Does Not Work On SLES11 SP1 with XEN (CR 6965290 and CR7110443)
I/O Does Not Work On SLES11 SP1 with XEN (CR 6965290
and CR7110443)
If the number of I/O devices exceeds XEN's number of interrupts, the I/O devices might not
work, and it might display corresponding error messages. For example:
Unable to allocate IRQ or map irq failed
Workaround
Add the following to the /boot/grub/menu.lst file:
extra_guest_irqs=64,2048 nr_irqs=2048
RHEL6 Kdump Runs Out of Memory (CR 7000993 and CR
7000942)
If Red Hat Linux 6 (RHEL6) calculates that the kdump size is too small, it displays a message:
"Your running kernel is using more than 70% of the amount of space you reserved
for kdump, you should consider increasing your crashkernel reservation"
One of two conditions might be present:
■
It might have enough memory, but post the messages anyway.
■
It might really not have enough kdump memory space. In this case, if the system crashes, it
might not produce a kdump file, and it might post messages similar to the following on the
console:
Out of memory: kill process 322 (insmod) score 21 or a child Killed process
322 (insmod) vsr:1352kB, anon-res: 244kB, file res: 122kB
Workaround
1. To make the messages stop and to ensure your system has enough kdump memory space,
change the following line in boot/grub/menu.lst:
Change:
crashkernel=auto
To:
Linux Issues53
Page 54
SLES11 SP1 Does Not Produce A Kdump File After A Crash (CR7001706)
crashkernel=512M
Note - This step stops your server from displaying warning messages. If your server does
not have enough kdump memory space, complete step 2 as well.
2.
Add pci=nomsi to the KDUMP_COMMANDLINE_APPEND line in /etc/sysconfig/kdump. For
example, change:
SLES11 SP1 Does Not Produce A Kdump File After A Crash
(CR7001706)
After a crash, SUSE Linux Enterprise Server 11 (SLES11) SP1 might not produce a kdump file.
In this case, it might also display a message similar to the following on the console:
Perform a regular update from Novell after installation.
SLES11 Fails to Install On Systems Equipped with 6–Core
CMODs (CR 7024769)
Servers equipped with 6-socket CMODs (Model E7540) might fail to install or boot with SUSE
Linux Enterprise Server 11 (SLES11).
Workaround
Install SLES11 SP1 or newer.
54Sun Fire X4800 Server Product Notes • July 2017
Page 55
Messages Complain of Failure to Allocate I/O Resources (CR 6984329)
Messages Complain of Failure to Allocate I/O Resources
(CR 6984329)
Messages that complain about failures to allocate I/O resources might appear in POST and in
log files.
For example, you might see:
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.0
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.1
Workaround
BIOS tries to allocate I/O resources. If it is unable to allocate I/O resources successfully, it
displays error messages.
The OS might try as well. However if it tries, it fails and generates additional failure messages
because it uses the same method as BIOS. However, most drivers can manage this condition.
1. Usually, you can ignore these messages.
2. If you continue to experience I/O resource issues, see I/O and Interrupt Resource Allocation
in the Sun Fire X4800 Server Installation Guide.
Xen Profiler Is Not Supported in Oracle Linux 5.5 (CR
6839366)
The use of the Xen Profiler is not supported in Oracle Linux 5.5.
Oracle Linux 5.5 Does Not Support CPU Throttling (CR
6847286)
CPU throttling (T-state) is not supported in Oracle Linux 5.5.
Linux Issues55
Page 56
InfiniBand PCIe Module Does Not Function Correctly With SLES 11 (CR 6909512, CR 6921163)
InfiniBand PCIe Module Does Not Function Correctly With
SLES 11 (CR 6909512, CR 6921163)
With SUSE Linux Enterprise Server 11 (SLES 11), if you use nox2apic, the InfiniBand PCIe
ExpressModule does not work with SLES 11.
Workaround
Install SLES11 SP1.
Bug in Kernel Causes Incorrect MSI Processing (CR
6921163)
If x2apic is enabled (if VT–d is enabled in BIOS), a bug in the kernel allows only a limited
number of interrupts to be used.
Workaround
Perform one or the other:
■
Update to the errata kernel from Novell.
■
Put nox2apic in the boot command line in /boot/grub/menu.1st append line.
LEDs on PCIe ExpressModule Work Incorrectly with Oracle
Linux 5.5 (CR 6894954)
The lights on the PCIe ExpressModule do not work normally on systems with Oracle Linux 5.5.
When you insert the PCIe ExpressModule and press the attention button:
■
The LED should blink five times for one second each time and then to stay on.
■
Instead, it stays dark for about five seconds and then stays on.
When you press the attention button to remove the PCIe ExpressModule:
■
The LED should blink for five seconds and then turn off.
56Sun Fire X4800 Server Product Notes • July 2017
Page 57
Oracle Linux 5.5 Xen Kernel Hangs Due to Long Timeout (CR 6899040)
■
Instead, it stays off for two seconds, blinks for three seconds, and then turns off.
Oracle Linux 5.5 Xen Kernel Hangs Due to Long Timeout
(CR 6899040)
When booting from a PCIe ExpressModule or a Fabric Expansion Module (FEM), the system
might hang because it runs out of MSI interrupt vectors.
Workaround
1.
In the GRUB configuration, set pci=nomsi.
2. Boot a non-Xen kernel.
3.
Edit the file /boot/grub/menu.lst.
4. Add the following text to the Xen kernel init line stanza:
pci=nomsi
For example:
# grub.conf generated by anaconda
#
# Note that you do not have to rerun grub after making changes to this file
# NOTICE: You have a /boot partition. This means that
# all kernel and initrd paths are relative to /boot/, eg.
10GbE PCIe Card in 8-Socket System with SLES11 Can
Cause Kernel to Crash (CR 6912821)
The kernel might crash in an 8-socket system with SUSE Linux Enterprise Server 11 (SLES11)
and a Sun Dual 10GbE I2 SFP+ PCIe 2.0. PCIe ExpressModule.
Workaround 1
Add the following entry to the /etc/modprobes.conf file:
options ixgbe InterruptType=1,1,1,1,1,1,1,1
It might be necessary to temporarily remove the Sun Dual 10GbE I2 SFP+ PCIe 2.0. PCIe
ExpressModule to prevent the kernel from crashing long enough for you to modify this file.
SLES11 with Xen Hangs When Many PCIe ExpressModule Cards Are Installed (CR 6917397)
SLES11 with Xen Hangs When Many PCIe ExpressModule
Cards Are Installed (CR 6917397)
SUSE Linux Enterprise Server 11 (SLES11) hangs with udev in dom0 when the number of
interrupt resources required by I/O devices exceeds a threshold. This is most likely in systems
where many PCIe ExpressModule cards are installed.
Workaround
1.
In the GRUB configuration, set pci=nomsi.
2. Boot a non-Xen kernel.
3.
Edit /boot/grub/menu.lst.
4. Add the following text to the Xen kernel init line stanza:
pci=nomsi
For example:
title Xen -- SUSE Linux Enterprise Server 11 - 2.6.27.19-5
“Messages Complain of Failure to Allocate I/O Resources (CR
6984329)” on page 61
“Earlier Versions of Oracle VM Only Support 64 Logical CPUs (CR
6915014)” on page 62
“Oracle VM Server Installation Parameters Required on Sun Fire X4800 (CR
7094126)” on page 62
Messages Complain of Failure to Allocate I/O Resources
(CR 6984329)
Messages that complain about failures to allocate I/O resources might appear in POST and in
log files.
Yes
Yes
Yes
For example, you might see:
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.0
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.1
Workaround
BIOS tries to allocate I/O resources. If it is unable to allocate I/O resources successfully, it
displays error messages.
The OS might try as well. However if it tries, it fails and generates additional failure messages
because it uses the same method as BIOS. However, most drivers can manage this condition.
1. Usually, you can ignore these messages.
Oracle VM Issues61
Page 62
Earlier Versions of Oracle VM Only Support 64 Logical CPUs (CR 6915014)
2. If you continue to experience I/O resource issues, see I/O and Interrupt Resource Allocation
in Sun Fire X4800 Server Installation Guide.
Earlier Versions of Oracle VM Only Support 64 Logical
CPUs (CR 6915014)
Earlier versions of Oracle VM (2.2.x) support 64 logical CPUs. An 8–socket system with
hyperthreading enabled has 128 logical CPUs.
Workaround
Disable hyperthreading in the BIOS.
Oracle VM Server Installation Parameters Required on Sun
Fire X4800 (CR 7094126)
To allow the megaraid_sas driver to load correctly you must add parameters to Oracle VM:
■
When installing Oracle VM 2.x, add the following kernel parameters:
If booting from the installation media, press F2 when the initial boot screen is displayed
and add the above parameters to the boot command.
■
If using a kickstart installation, add the above parameters to the PXE configuration file.
■
For all Oracle VM installations, edit the /boot/grub/grub.conf file in your Oracle VM
Server and add: extra_guest_irqs=64,2048 nr_irqs=2048
62Sun Fire X4800 Server Product Notes • July 2017
Page 63
Windows Operating System Issues
The following table lists the Windows OS issues.
Windows IssuesWorkaround
“Unspecified CPU Fault After Warm Reset (CR 7054657)” on page 63Yes
“Hot Plugging of Some PCIe ExpressModules Not Supported by Windows Server
2008 R2 (CR 6914898, CR 6793369, CR 6923132)” on page 64
“PXE Installation of Windows Server 2008 R2 Fails at Authentication (CR
6899582)” on page 64
“Messages Complain of Failure to Allocate I/O Resources (CR
6984329)” on page 65
“Windows 2008 Cannot Be Installed to Combo GbE/8Gb FC ExpressModule HBA
Disks (CR 6984746)” on page 66
“Hot Plugging of PCIe ExpressModules Is Not Supported by Windows
2008” on page 66
Unspecified CPU Fault After Warm Reset (CR 7054657)
Yes
Yes
Yes
Yes
No
When the server undergoes a reset because of a CPU accessing memory from an uncorrectable
DIMM, a reboot or warm reset from the server might result in an unspecified CPU fault. This
has been observed with Windows 2008 R2 SP1 and Windows 2008 SP2.
The sequence of events is:
■
CPU accesses memory from an uncorrectable ECC DIMM.
■
As a result, the CPU initiates a platform reset.
■
The next reset results in a valid DIMM fault.
■
When the DIMMs are not replaced and the system continues to operate with the fault,
a subsequent reboot or warm reset results in processors faults. However, these are false
processor faults
Workaround
Windows Operating System Issues63
Page 64
Hot Plugging of Some PCIe ExpressModules Not Supported by Windows Server 2008 R2 (CR 6914898, CR 6793369, CR
6923132)
1. Ignore these false processor faults when they're associated with a uncorrectable DIMM
reset.
2. Handle the DIMM errors like you would any other DIMM error, and clear the DIMM faults.
3. Clear the processor faults.
4. Power-cycle the server.
For example, in the Oracle ILOM Remote Power Control tab, select Power Cycle or
Immediate Power Off.
Note - Resetting the server without toggling the power (for example, by issuing a reset
command) does not clear the CPU error.
Hot Plugging of Some PCIe ExpressModules Not Supported
by Windows Server 2008 R2 (CR 6914898, CR 6793369, CR
6923132)
Windows Server 2008 R2 does not support hot plugging on certain PCIe ExpressModules.
These include:
■
SG-XPCIE2FCGBE-Q-Z - Sun StorageTek Dual FC Dual GbE HBA Express Module
■
SG-XPCIE2FCGBE-E-Z - Sun StorageTek Dual FC Dual GbE HBA Express Module
■
SG-XPCIEFCGBE-Q8-Z - Sun StorageTek Dual 8GB FC Dual GbE HBA Express Module
■
SG-XPCIEFCGBE-E8-Z - Sun StorageTek Dual 8GB FC Dual GbE HBA Express Module
■
X7284A-Z - Sun Quad GbE ExpressModule
Workaround
Reboot the system after changing one of these PCIe ExpressModules.
PXE Installation of Windows Server 2008 R2 Fails at
Authentication (CR 6899582)
During a PXE installation of Windows Server 2008 R2, if PXE user authentication fails, check
that the system date and time are set correctly. If the systems date and time is incorrect, PXE
authentication fails, and the network installation of Windows Server 2008 R2 does not happen.
64Sun Fire X4800 Server Product Notes • July 2017
Page 65
Messages Complain of Failure to Allocate I/O Resources (CR 6984329)
Note - This failure occurs when the date is approximately seven days different from the actual
date.
Workaround
Use the system BIOS to set the system date and time.
1. Reboot or power on the server.
POST messages appear, and prompt you to press the F2 key.
2. Press the F2 key.
The BIOS Setup Utility main menu appears.
■
Use the left and right arrow keys or the Tab key to navigate through the top level of the
utility.
■
Use the up and down arrows to navigate through the submenus.
■
Use the Help information for instructions to change options, save settings, and exit the
utility.
3. Find the system date and time and set it.
4. Save your work and exit.
Messages Complain of Failure to Allocate I/O Resources
(CR 6984329)
Messages that complain about failures to allocate I/O resources might appear in POST, and in
log files.
For example, you might see:
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.0
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.1
Workaround
BIOS tries to allocate I/O resources. If it is unable to allocate I/O resources successfully, it
displays error messages.
The OS might try as well. However if it tries, it fails and generates additional failure messages
because it uses the same method as BIOS. However, most drivers can manage this condition.
1. Usually, you can ignore these messages.
Windows Operating System Issues65
Page 66
Windows 2008 Cannot Be Installed to Combo GbE/8Gb FC ExpressModule HBA Disks (CR 6984746)
2. If you continue to experience I/O resource issues, see I/O and Interrupt Resource Allocation
in Sun Fire X4800 Server Installation Guide.
Windows 2008 Cannot Be Installed to Combo GbE/8Gb FC
ExpressModule HBA Disks (CR 6984746)
If you attempt to install Windows 2008 on a Combo GbE/8Gb FC ExpressModule HBA Disk,
the driver loads but the OS cannot see the disks.
Workaround
During the installation, load the drivers from the USB drive twice.
Loading the drivers twice is necessary because there are two devices on the Combo GbE/8Gb
FC ExpressModule HBA.
Hot Plugging of PCIe ExpressModules Is Not Supported by
Windows 2008
With Windows 2008 SP2 and 2008 R2 the following PCIe ExpressModules cannot be hotplugged:
“Messages Complain of Failure to Allocate I/O Resources (CR
6984329)” on page 67
“Option Cards Can Fail to Load Device Drivers in Some Configurations (CR
6933436)” on page 68
“Warning Messages Appear During Bootup and in Log Files (CRs 6862766,
6861903, and 6890263)” on page 69
“VMware ESXi 5.5 Does Not Support MMIO Regions Above 4GB (16480679,
17013064)” on page 69
“VMware ESXi 5.5 Runs Out of Interrupts With PCIe Cards
(16494653)” on page 70
Messages Complain of Failure to Allocate I/O Resources
(CR 6984329)
Yes
Yes
N/A
Yes
Yes
Messages that complain about failures to allocate I/O resources might appear in POST, and in
log files.
For example, you might see:
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.0
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.1
Workaround
BIOS tries to allocate I/O resources. If it is unable to allocate I/O resources successfully, it
displays error messages.
ESX Issues67
Page 68
Option Cards Can Fail to Load Device Drivers in Some Configurations (CR 6933436)
The OS might try as well. However if it tries, it fails and generates additional failure messages
because it uses the same method as BIOS. However, most drivers can manage this condition.
1. Usually, you can ignore these messages.
2. If you continue to experience I/O resource issues, see I/O and Interrupt Resource Allocation
in Sun Fire X4800 Server Installation Guide.
Option Cards Can Fail to Load Device Drivers in Some
Configurations (CR 6933436)
Some combinations of option cards can result in failure to load device drivers and subsequent
failure to mount file systems or configure network interfaces, and possibly failure to boot the
system. This occurs in ESX 4x only.
The ESX host requires "Low Memory," or the memory addressed below 4 GB for the Service
Console, device drivers and other data structures. The BIOS utilizes 2 GB of this memory area
for its own uses. With the large number of supported I/O devices, this memory can become
exhausted in some situations.
VMware describes this problem in their Knowledge Base at https://vmware.com.
Workaround
Use the following procedure to reduce the amount of memory reserved by the Service Console.
1. Restart the ESX host.
2. When the boot menu appears, ensure that the default boot option is highlighted.
3. Press e to edit the default boot option.
A new list appears.
4.
Use the arrow keys to highlight the line that starts with kernel /vmlinuz.
5. Press the e key to edit the line in the boot configuration.
6. Use the arrow key to move over the portion of the line which has a parameter similar to the
following:
mem=800M
7. Change the value to 600 megabytes.
For example:
mem=600M
8. Press Enter to accept the changes.
9. Press the b key to boot the host using the new configuration.
68Sun Fire X4800 Server Product Notes • July 2017
Page 69
Warning Messages Appear During Bootup and in Log Files (CRs 6862766, 6861903, and 6890263)
The configuration is complete but it is not yet permanent.
10. After the host has booted, open the vSphere Client and connect to the host.
11. Select the Configuration tab.
12. Change the memory properties of the ESX host to permanently reduce the RAM
assignment.
Warning Messages Appear During Bootup and in Log Files
(CRs 6862766, 6861903, and 6890263)
Various warning messages appear during bootup, and in log files such as /var/log/
vmkwarning. These can be ignored, as they have no impact on performance.
Sample messages include:
Sep 24 02:05:07 mpk12-3214-189-114 vmkernel: 0:00:00:24.984
cpu0:4096: WARNING PCI: 1861: No such device
Sep 24 02:05:07 mpk12-3214-189-114 last message repeated 9 times
Sep 24 02:05:07 mpk12-3214-189-114 vmkernel: 0:00:01:03.280 cpu49:4166)WARNING:
vmklinux26: dma_alloc_coherent: Out of memory
Sep 24 04:13:21 mpk12-3214-189-114 vmkernel: 0:00:00:38.761 cpu0:4096)WARNING:
PCI: 1861: No such device
Sep 24 04:13:21 mpk12-3214-189-114 last message repeated 9 times
Sep 24 04:13:22 mpk12-3214-189-114 vmkernel: 0:00:01:19.828 cpu16:4166)WARNING:
vmklinux26: dma_alloc_coherent: Out of memory
VMware ESXi 5.5 Does Not Support MMIO Regions Above
4GB (16480679, 17013064)
The Sun Server X4800 defaults in BIOS to 64-bit MMIO (Memory Mapped I/O). This allows
additional PCIe memory address space to be mapped above the standard 32-bit 4GB of space
for PCIe cards that include option ROMs. However, VMware ESXi is incompatible with MMIO
space above the standard 4GB. This issue can cause some PCIe cards not to function properly
with ESXi.
Workaround
As a possible workaround, disable 64-bit MMIO through the server's BIOS Setup utility (under
the IO > PCIe Subsystem Settings > PCI 64-bit Resources Allocation menu). This workaround
has limitations. With some combinations of option cards, the system will require more MMIO
ESX Issues69
Page 70
VMware ESXi 5.5 Runs Out of Interrupts With PCIe Cards (16494653)
space than what the system can allocate within 32 bits of address space. When that occurs, those
option cards that could not be assigned MMIO address space (because there was not enough
left) are unavailable for use.
VMware ESXi 5.5 Runs Out of Interrupts With PCIe Cards
(16494653)
In certain configurations VMware ESXi can run out of interrupts for devices (this can include
storage and networking).
For more information, refer to VMware's Configuration Maximums document for ESXi 5.5,
under host maximums:
Note - Because of Java security changes, with the previous Oracle ILOM firmware if you used
Java 7 update 51 or newer, you could not launch Oracle ILOM Remote Console. With the new
Oracle ILOM firmware if use Java 7 update 51 or newer, you can launch Oracle ILOM Remote
Console.
Oracle ILOM IssuesWorkaround
“File Transfer Using URI Fails if Target Password Contains Certain Special
Characters (25917655)” on page 72
“Oracle ILOM SNMP v3 Traps Are Not Delivered After SNMP Engine ID Change
(23634048)” on page 72
“Third-Party Web Scan and Test Tools Cause Sluggish Oracle ILOM Performance
(23564626)” on page 73
“Oracle ILOM Enhancement Allows IPv4 Only, IPv6 Only, or Dual
Stack” on page 74
“New Procedures for Updating CPLD (CR 7043418)” on page 75N/A
“Update Oracle ILOM and BIOS Firmware Before Updating Other Device Firmware
(CR 6537282)” on page 75
“Use the Locate Button to Prove Physical Presence (CR 6881237)” on page 76N/A
“NEM Expander Firmware Update Procedure (CR 6979140)” on page 76Yes
“Network Management Port 1 Does Not Work (CR 6935471 and
6854694)” on page 81
“start /SYS and stop /SYS Commands Cause Power Button Pressed Event In Log
(CR 6906176)” on page 81
“Event Log Messages Might Be Out of Sequence and Duplicated (CR
6915976)” on page 81
“Incorrect Error Message in Event Log After Restore When Serial Console or
JavaRconsole Session Are Open (CR 6917474)” on page 82
“Allocated Power Figures Are Incomplete (CR 6931837)” on page 82No
Fixed in SW 1.1
“(Fixed in SW 1.1) LDAP Account Cannot Be Used to Start Console (CR
6969473)” on page 83
Fixed in System Software
Release 1.11.0
Yes
Yes
N/A
N/A
No
No
No
No
Yes
Oracle ILOM Enhancements and Issues71
Page 72
File Transfer Using URI Fails if Target Password Contains Certain Special Characters (25917655)
Oracle ILOM IssuesWorkaround
“(Fixed in SW 1.1) Power Cycling the Host Using the Web Interface Generates an
Error (CR 6909374)” on page 83
“(Fixed in SW 1.1) DIMM Mismatch Fault Not Cleared After DIMM Replacement
and Host Power Cycle (CR 6972285)” on page 83
Fixed in SW 1.0.1
“(Fixed in SW 1.0.1) Console Redirect from CLI Fails (CR 6939479)” on page 84Yes
“(Fixed in SW 1.0.1) Oracle ILOM Only Allows One Console Session (CR
6895423)” on page 84
“(Fixed in SW 1.0.1) Faults Disappear After SP Reset (CR 6928998)” on page 85Yes
“(Fixed in SW 1.0.1) Clearing a Power Supply Unit (PSU) Fault Erases Other Fault
Information (CR 6933348 and CR 6939306)” on page 85
“(Fixed in SW 1.0.1) pciflash Is Not Available (CR 6867060 and CR
6904922)” on page 85
No
Yes
No
Yes
Yes
File Transfer Using URI Fails if Target Password Contains
Certain Special Characters (25917655)
This problem is fixed in System Software Release 1.11.0.
When using Oracle ILOM to transfer files using a Uniform Resource Identifier (URI), the
transfer fails if the target host's password contains any of the following special characters:
# ; ?
Examples of these transfers include using host storage redirection, and backing up and restoring
BIOS and SP configurations.
Workaround
Use a target host password that does not include any of the indicated special characters.
Oracle ILOM SNMP v3 Traps Are Not Delivered After SNMP
Engine ID Change (23634048)
If you change the engine ID, create an SNMP v3 user, and configure an alert using that user
without waiting approximately 10 seconds between each action, the internal user configuration
might be incorrect and traps are missed.
72Sun Fire X4800 Server Product Notes • July 2017
Page 73
Third-Party Web Scan and Test Tools Cause Sluggish Oracle ILOM Performance (23564626)
Workaround
Do not create multiple configuration changes without verifying the effect of each configuration
change. To prevent misconfigured users and missed traps, insert sleep statements in the script.
For example:
# change engineID
set /SP/services/snmp engineid=NEWENGINEID
# sleep 10 seconds to give snmp enough time to make the change
sleep 10
# verify engineID
show /SP/services/snmp engineid
# verify SNMPv3 users have been deleted
show /SP/services/snmp/users
# create snmpv3 user
create /SP/services/snmp/users newuser authenticationpassword=...
# sleep 10 seconds to give snmp enough time to make the change
sleep 10
# verify user
show /SP/services/snmp/users newuser
# do a snmpget with that user to verify it
# configure alert
set /SP/alertmgmt/rules/1 type=snmptrap ...
# sleep 10 seconds to give snmp enough time to make the change
sleep 10
# verify alert
show /SP/alertmgmt/rules/1
set /SP/alertmgmt/rules/1 testrule=true
Third-Party Web Scan and Test Tools Cause Sluggish
Oracle ILOM Performance (23564626)
Under certain conditions, third-party web scanning and test tools can cause Oracle ILOM to run
extremely slowly.
On systems with system software release 1.6 on newer, Oracle ILOM supports the ability
to independently enable or disable IPv4 and IPv6. Also you can now configure a static IPv6
gateway.
1.
Log in to Oracle ILOM as an Administrator.
For instructions on how launch Oracle ILOM from the CLI or web interface, see: Connecting to
the ILOM in Sun Fire X4800 Server Installation Guide.
2.
To modify the SP network settings, perform one of the following:
Modify the settings on the Network Settings page as required.
For further details about how to configure the properties on the Network Setting page,
click the More Details link.
c.
Click Save to save your network property changes in Oracle ILOM.
Note - When you save your network settings, it might end your Oracle ILOM session. If this
happens, use the new settings to log in again.
■From the CLI interface, perform these steps:
a.
To view the assigned IPv4 and IPv6 network settings on the SP, perform
the following:
For IPv4, type: show /SP/network
For IPv6, type: show /SP/network/ipv6
b.
To view the descriptions about each IPv4 and IPv6 network property,
perform the following:
For IPv4, type: help /SP/network
For IPv6, type: help /SP/network/ipv6
74Sun Fire X4800 Server Product Notes • July 2017
Page 75
New Procedures for Updating CPLD (CR 7043418)
c.
To modify the IPv4 and IPv6 network properties on the SP, issue the set
command.
IPv4 Example:
set /SP/network state=enabled|disabled pendingipdiscovery=static|dhcp
pendingipaddress=value pendingipgateway=value pendingipnetmask=value
IPv6 Example:
set /SP/network/ipv6 state=enabled|disabled pending_static_ipaddress=value/
subnet_mask_value pending_static_ipgatewayaddress=value
Note - The set /SP/network/state = enabled command does not enable IPv6. To enable
IPv6, use either set /SP/network/ipv6/state = enabled or (with SW 1.6 or newer), use
set /SP/network/state = ipv6-only.
■To commit the IPv4 and IPv6 pending network changes, type: set /SP/network
commitpending=true
Note - If you are logged in to Oracle ILOM using an Ethernet connection, your connection is
terminated when you set commitpending to true. When this happens, log back in using the new
settings.
New Procedures for Updating CPLD (CR 7043418)
The procedures for updating CPLD have changed. For details, refer to Updating the CPLD in
Sun Fire X4800 Server Service Manual.
Update Oracle ILOM and BIOS Firmware Before Updating
Other Device Firmware (CR 6537282)
When updating the firmware for multiple devices, if your list of firmware updates includes
Oracle ILOM and BIOS firmware, update it first.
Oracle ILOM Enhancements and Issues75
Page 76
Use the Locate Button to Prove Physical Presence (CR 6881237)
Use the Locate Button to Prove Physical Presence (CR
6881237)
Certain Oracle ILOM procedures require you to “prove physical presence”. To do this, press the
Locate button located on the system front panel.
For details, see Front and Back Features and Components in Sun Fire X4800 ServerInstallation Guide.
NEM Expander Firmware Update Procedure (CR 6979140)
Problems with the NEM firmware can cause various symptoms, including but not limited to:
■
System won't boot and displays an fma fault:
sunw-msg-id #SPX86–8002–QQID
■
Partial or no information in sensor data lists, especially in NEM sensors.
To fix these problems, you must update the NEM expander firmware.
Before You Begin
1.
■
“Workaround 1 Using Oracle ILOM Web GUI” on page 76
■
“Workaround 2 Using Oracle ILOM CLI” on page 79
Workaround 1 Using Oracle ILOM Web GUI
Perform the following actions before upgrading your NEM firmware:
■
Download the firmware image for your server or CMM from the platform's product web
site. For details, see https://support.oracle.com.
■
Copy the firmware image to the system on which the web browser is running, using a
supported protocol (TFTP, FTP, HTTP, HTTPS).
■
Obtain an Oracle ILOM user name and password that has Administrator role account
privileges. You must have Administrator privileges to update the firmware on the system.
Restart the server and enter the BIOS screen.
When POST messages appear, press F2 to enter the BIOS Setup Utility.
Note - You do not need to configure anything in the BIOS Setup Utility. It is used in this
procedure to ensure that the NEMs are powered-on but the OS does not boot.
76Sun Fire X4800 Server Product Notes • July 2017
Page 77
Workaround 1 Using Oracle ILOM Web GUI
2.
Log in to the Oracle ILOM web interface.
For more details, see How to Connect to the ILOM Web Interface in Sun Fire X4800 Server
Installation Guide.
3.
Select System Information –> Components.
The Component Management page appears.
4.
Highlight NEM0.
5.
From the Actions drop-down menu, select Update Firmware.
A screen asks for download details.
Oracle ILOM Enhancements and Issues77
Page 78
Workaround 1 Using Oracle ILOM Web GUI
Note - If the Current Version is FW 5.3.3.4 or greater, you do not need to update your firmware.
6.
Perform the following actions:
a.
Specify the image location by performing one of the following:
■Click Browse to select the location of the firmware image you want to
install.
■If supported on your system, click Specify URL. Then type the URL for
the firmware image into the text box.
b.
Select a transfer method from the drop-down list.
c.
Click the Update button to upload the file and update the firmware.
The Update Status display appears, providing details about the update progress. When the
update indicates 100%, the firmware upload is complete.
When the update is finished, it displays the message Firmware Update Successful.Auto-closing windows in 2 seconds, then it closes the display.
7.
Repeat steps 1 through 6 for NEM1.
78Sun Fire X4800 Server Product Notes • July 2017
Page 79
8.
Press ESC in the BIOS Setup Utility to boot the system.
Workaround 2 Using Oracle ILOM CLI
Workaround 2 Using Oracle ILOM CLI
Before You Begin
1.
2.
3.
4.
Perform the following actions before upgrading your NEM firmware:
■
Download the firmware image for your server or CMM from the platform's product web
site. For details, see https://support.oracle.com.
■
Copy the firmware image to the system on which the web browser is running, using a
supported protocol (TFTP, FTP, HTTP, HTTPS).
■
Obtain an Oracle ILOM user name and password that has Administrator role account
privileges. You must have Administrator privileges to update the firmware on the system.
Restart the server and enter the BIOS screen.
When POST messages appear, press F2 to enter the BIOS the BIOS Setup Utility.
Note - You do not need to configure anything in the BIOS the BIOS Setup Utility. It is used in
this procedure to ensure that the NEMs are powered-on but the OS does not boot.
Log in to the Oracle ILOM CMM CLI.
Use the cd command to navigate to NEM0.
Type the show command to see the NEM properties, including the firmware
version.
The fru_extra_1 property field identifies the firmware version presently installed on the NEM.
If it is 5.3.3.4 or greater, you do not need to update your NEM firmware.
-> cd /SYS/NEM0
/SYS/NEM0
-> show
/SYS/NEM0
Targets:
PRSNT
STATE
ERR
OK
SERVICE
OK2RM
LOCATE
ip_address is the IP address of the system where the file is stored.
■
username is the login user name to the system where the file is stored.
■
password is the login password to the system where the file is stored.
■
rom_nem.pkg is the name of the firmware update package.
6.
Wait a few moments for Oracle ILOM to confirm the completion of the firmware
update process.
A success or failure status appears.
80Sun Fire X4800 Server Product Notes • July 2017
Page 81
Network Management Port 1 Does Not Work (CR 6935471 and 6854694)
7.
Use the show command to view and confirm the firmware version that is installed
on the NEM.
8.
Repeat steps 1 through 7 for NEM1.
9.
Press ESC in the BIOS Setup Utility to restart the boot process.
Network Management Port 1 Does Not Work (CR 6935471
and 6854694)
The CMM has two network management ports. The connectors are labelled NET MGT 0 and
NET MGT 1.
NET MGT 1 does not work. Use NET MGT 0 to connect the CMM to the Ethernet.
If you need to connect to more than one network, use an external Ethernet switch.
Note - Oracle ILOM allows you to select which management port to use. Even if you select port
1, it does not switch ports.
start /SYS and stop /SYS Commands Cause Power Button
Pressed Event In Log (CR 6906176)
When you enter the start /SYS and stop /SYS commands, it causes a Power Button Pressed
event to be logged. This log entry is incorrect.
You can ignore these log entries.
Event Log Messages Might Be Out of Sequence and
Duplicated (CR 6915976)
Messages in the event log might be out of sequence and duplicated. For example:
14 Fri Apr 16 05:42:50 2010 IPMI Log critical
ID = a : 04/16/2010 : 05:42:50 : System Firmware Progress : BIOS : Man
agement controller initialization
Oracle ILOM Enhancements and Issues81
Page 82
Incorrect Error Message in Event Log After Restore When Serial Console or JavaRconsole Session Are Open (CR 6917474)
13 Fri Apr 16 05:42:49 2010 IPMI Log critical
ID = 9 : 04/16/2010 : 05:42:49 : System Boot Initiated : BIOS : Initia
ted by hard reset
12 Fri Apr 16 05:42:48 2010 IPMI Log critical
ID = a : 04/16/2010 : 05:43:54 : Button : sensor number = 0x00 : Power
Button pressed
11 Fri Apr 16 05:42:47 2010 IPMI Log critical
ID = 8 : 04/16/2010 : 05:42:47 : Module/Board : BL3/STATE : Transition
to Running
10 Fri Apr 16 05:35:36 2010 IPMI Log critical
ID = 7 : 04/16/2010 : 05:35:36 : Entity Presence : BL0/FMOD1/PRSNT : D
evice Absent
9 Fri Apr 16 05:35:27 2010 IPMI Log critical
ID = 6 : 04/16/2010 : 05:35:27 : Module/Board : BL0/STATE : Transition
to Power Off
Note message number 12; its ID should be 8, and the IDs of subsequent messages should be
decremented accordingly.
Incorrect Error Message in Event Log After Restore When
Serial Console or JavaRconsole Session Are Open (CR
6917474)
If you do a restore while a serial console or a JavaRconsole session are open, you might see an
error message in the event log. For example:
409 Restore Log major Fri Feb 26 19:42:40 2010
Config restore: Unable to restore property '/SP/serial/host/commitpending'.
This message is incorrect. You can ignore it.
Workaround
If a JavaRconsole session or a serial console is open, use the CLI to set console properties
manually. Do not use a configuration restore.
Allocated Power Figures Are Incomplete (CR 6931837)
When you select Power Management – > Allocation in the web interface it shows the power
consumption numbers for various server components. It also shows the total consumed by the
server.
82Sun Fire X4800 Server Product Notes • July 2017
Page 83
(Fixed in SW 1.1) LDAP Account Cannot Be Used to Start Console (CR 6969473)
There are two anomalies in the displayed results:
■
The display does not show all the components in the server. For example, it does not show
processor modules.
■
As a result, the sum used by the (displayed) components is less than the total used by the
system.
This is not really a problem. The Allocated Power figure shows how much power the server is
using.
(Fixed in SW 1.1) LDAP Account Cannot Be Used to Start
Console (CR 6969473)
If you use an LDAP account to log into Oracle ILOM, the start /SP/console command does
not work.
Workaround
Log into Oracle ILOM using a non-LDAP account if you want to use the console.
(Fixed in SW 1.1) Power Cycling the Host Using the Web
Interface Generates an Error (CR 6909374)
If you use the Oracle ILOM web interface to power cycle the server, it might display an error
message stating that the operation failed, even though it succeeded.
You can ignore this error message.
(Fixed in SW 1.1) DIMM Mismatch Fault Not Cleared After
DIMM Replacement and Host Power Cycle (CR 6972285)
If you have a DIMM mismatch fault and you replace the faulted DIMM and restart the host, the
fault is not cleared.
Workaround
Oracle ILOM Enhancements and Issues83
Page 84
(Fixed in SW 1.0.1) Console Redirect from CLI Fails (CR 6939479)
(Fixed in SW 1.0.1) Console Redirect from CLI Fails (CR
6939479)
Under rare conditions, attempting to start the console using the CLI command start /SP/
console fails.
When this happens, Oracle ILOM returns to the CLI prompt without displaying any warning
messages.
When the console redirection command is successful, it looks like this:
-> start /SP/console
Are you sure you want to start /SP/console (y/n) y
Disabling external host serial connection
Serial console started. To stop, type ESC (
When the command fails, it looks like this:
-> start /SP/console
Are you sure you want to start /SP/console (y/n) y
->
Workaround
1. Reboot the SP.
Use the command reset SP.
2. Wait for the SP to fully reboot.
3.
Try the start /SP/console command again
The console redirect works after the SP boots.
(Fixed in SW 1.0.1) Oracle ILOM Only Allows One Console
Session (CR 6895423)
Oracle ILOM only supports a single console session, which is started using the start /SP
console command.
84Sun Fire X4800 Server Product Notes • July 2017
Page 85
(Fixed in SW 1.0.1) Faults Disappear After SP Reset (CR 6928998)
(Fixed in SW 1.0.1) Faults Disappear After SP Reset (CR
6928998)
Resetting the SP clears information about faults, but does not remove the actual faults.
Workaround
Before you reset the SP, enter the show faulty command and save the output to retain a record
about existing faults.
(Fixed in SW 1.0.1) Clearing a Power Supply Unit (PSU)
Fault Erases Other Fault Information (CR 6933348 and CR
6939306)
When the service processor clears a power supply fault, it also erases other fault records. This
might happen:
■
When you hotswap a power supply. However, you can preserve the information by creating
copies of the other faults.
■
Occasionally a power supply fault clears itself. When this happens, the information about
other faults might be lost.
Workaround
Perform this procedure before hotswapping a power supply.
1. Save your fault information immediately when you see a power fault, and before hot
swapping a power supply.
■
In the web interface, this information is located in the Fault Management tab.
■
In the CLI, this information is in fmadm.
(Fixed in SW 1.0.1) pciflash Is Not Available (CR 6867060
and CR 6904922)
The command pciflash is not available. As a result, the command ipmiflash —l pci does not
work either.
Oracle ILOM Enhancements and Issues85
Page 86
(Fixed in SW 1.0.1) pciflash Is Not Available (CR 6867060 and CR 6904922)
Workaround
Use other methods to update Oracle ILOM, as described in the Oracle Integrated Lights Out
Manager (ILOM) 3.0 Supplement for the Sun Fire X4800 Server.
86Sun Fire X4800 Server Product Notes • July 2017
Page 87
Oracle Hardware Installation Assistant Issues
The following table lists the Oracle Hardware Installation Assistant issues.
Oracle Hardware Installation Assistant Image Does Not
Boot On a System With More Than 512 GB of Memory (CR
6972927)
When you boot Oracle Hardware Installation Assistant on a system with more than 512 GB of
memory, the system repeatedly resets itself.
Oracle Hardware Installation Assistant Issues87
Page 88
88Sun Fire X4800 Server Product Notes • July 2017
Page 89
BIOS Issues
The following table lists the BIOS issues for Oracle's Sun Fire X4800 server.
BIOS IssuesWorkaround
“Auto Boot Host On Power Loss Control Is Deactivated” on page 89N/A
“Update Oracle ILOM and BIOS Firmware Before Updating Other Devices' Firmware (CR
6537282)” on page 89
“Hotplugging PCIe EM Card Fails When Replacing the Original Card With One of a Different
Type (CR 7003634)” on page 90
“Messages Complain of Failure to Allocate I/O Resources (CR 6984329)” on page 90Yes
“DIMM Failure Causes Other DIMMs To Be Disabled (CR 6929978)” on page 91N/A
Auto Boot Host On Power Loss Control Is Deactivated
The control to Auto Boot Host On Power Loss in the BIOS Setup Utility has been deactivated.
N/A
Yes
To set Auto Boot Host On Power Loss, use Oracle ILOM instead.
Update Oracle ILOM and BIOS Firmware Before Updating
Other Devices' Firmware (CR 6537282)
When updating the firmware for multiple devices, if your list of devices includes Oracle ILOM
and BIOS firmware, update it first.
BIOS Issues89
Page 90
Hotplugging PCIe EM Card Fails When Replacing the Original Card With One of a Different Type (CR 7003634)
Hotplugging PCIe EM Card Fails When Replacing the
Original Card With One of a Different Type (CR 7003634)
When you remove a PCIe ExpressModule (EM) card and replace it by hotplugging a card of
a different type, the OS might not be able to configure the new card. This is because before
booting to the OS, the BIOS allocates resources based on the card's requirements during POST.
If the new card requires more resources than the original, the OS does not have the ability to
allocate more resources to the new card at this time, and the card does not operate correctly.
Workaround 1
Hotplug PCIe EM cards in previously empty slots.
Workaround 2
If a hotplugged PCIe EM card fails to operate correctly, reboot the server.
Messages Complain of Failure to Allocate I/O Resources
(CR 6984329)
Messages that complain about failures to allocate I/O resources might appear in POST, and in
log files.
For example, you might see:
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.0
Sep 8 15:50:49 nsg14-28 kernel: PCI: Failed to allocate I/O resource
#2:20@0 for 0000:8d:00.1
Workaround
BIOS tries to allocate I/O resources. If it is unable to allocate I/O resources successfully, it
displays error messages.
The OS might try as well. However if it tries, it fails and generates additional failure messages
because it uses the same method as BIOS. However, most drivers can manage this condition.
Usually, you can ignore these messages.
If you continue to experience I/O resource issues, see I/O and Interrupt Resource Allocation in
Sun Fire X4800 Server Installation Guide.
90Sun Fire X4800 Server Product Notes • July 2017
Page 91
DIMM Failure Causes Other DIMMs To Be Disabled (CR 6929978)
DIMM Failure Causes Other DIMMs To Be Disabled (CR
6929978)
When a pair of DIMMs is disabled due to faults, BIOS disables other DIMMs to maintain a
valid configuration. For example:
■
In an 8-socket system, if a single DIMM is disabled due to faults, BIOS disables either 4 or
8 DIMMs depending on the socket number:
Faulted DIMMDisabled DIMMs
D0 or D4D0, D4, D1, D5, D8, D9, D12, D13
D1 or D5D1, D5, D9, D13
D2 or D6D2, D6, D3, D7, D10, D11, D14, D15
D3 or D7D3, D7, D11, D15
D8 or D12D0, D4, D1, D5, D8, D9, D12, D13
D9 or D13D1, D5, D9, D13
D10 or D14D2, D6, D3, D7, D10, D11, D14, D15
D11 or D15D3, D7, D11, D15
■
In a 4–socket system, a faulted DIMM causes the BIOS to disable either two or four
DIMMs, depending on the socket number.
Faulted DIMMDisabled DIMMs
D0 or D4D0, D4, D1, D5
D1 or D5D1, D5
D2 or D6D2, D6, D3, D7
D3 or D7D3, D7
D8 or D12D8, D9, D12, D13
D9 or D13D9, D13
D10 or D14D10, D11, D14, D15
D11 or D15D11, D15
Note - When faulty happens on odd-numbered DIMMS, see “DIMM Mismatch on Odd
Numbered Pairs Results in DDR Training Failed on Entire Branch (CR7101624 and
CR7111545)” on page 34.
Workaround
1. Identify the faulted DIMMs.
BIOS Issues91
Page 92
DIMM Failure Causes Other DIMMs To Be Disabled (CR 6929978)
Open the Oracle ILOM fault management tab.
Note - DIMMs are always faulted in pairs, and should always be replaced in pairs.
■
Faulty DIMMs are listed as:
■
fault.memory.intel.nex.dimm_ce, or
■
fault.memory.intel.nex.dimm_ue
■
In 4–socket systems, Oracle ILOM only lists the faulted DIMMs. In 8–socket systems,
Oracle ILOM also lists disabled DIMMs. These are listed as fault.memory.intel.dimm.population.invalid.
■
If all the DIMM are listed as fault.memory.intel.dimm.population.invalid, then
the configuration is invalid.
2. Replace the faulted DIMMs.
Refer to Removing and Installing DIMMs (CRU) in Sun Fire X4800 Server Service Manual.
3. Clear the DIMM faults in Oracle ILOM.
Refer to your Oracle ILOM documentation collection for details.
92Sun Fire X4800 Server Product Notes • July 2017
Loading...
+ hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.