Sun Fire X4600 M2 Server Product Notes

This chapter describes the LSI firmware, Sun Integrated Lights Out Manager (ILOM) service processor (SP), and BIOS issues related to the Sun Fire X4600 M2 server. The numbers given in the section titles are internal tracking numbers for change requests related to the issues. These issues are divided into the following topics:

• Service Processor Firmware Issues
• Resolved Service Processor Issues
• BIOS Issues
• Resolved BIOS Issues
• LSI Firmware Issues

---

Service Processor Firmware Issues

The following are current issues.

Mouse Pointers Out of Sync on Windows 2003 at High Resolution (6669368)

Special Character in FTP Password Disables Snapshot Feature (6674741)

Service Processor Does Not Log Event or Provide Visual Alert After Hard Disk Is Removed

Recovery For Accidentally Flashing Wrong Platform Firmware on Sun Fire X4600 M2 Server

Use -f Option To Upgrade ILOM when Updating Firmware

When you upgrade the system firmware, use the load command with the -f option. This upgrades the ILOM as well.

-> load -f other_options

See the ILOM documentation collection for additional details.

OEM Error Messages In ILOM Logs (6608347)

If you see OEM error messages in the event log, contact Oracle Customer Support.

The following table shows an example of OEM error messages in the event log:

7b02 | 09/13/2007 | 16:07:27 | OEM #0x12 |  | Asserted
7c02 | OEM record e0 | 00004800001111032000000000
7d02 | OEM record e0 | 10ab0000000810000006040012
7e02 | OEM record e0 | 10ab0000001111002011110020 

Mouse Pointers Out of Sync on Windows 2003 at High Resolution (6669368)

When JavaRConsole is used to access Windows 2003, and the remote display resolution is greater than 1024x768, the local mouse pointer goes out of sync with the remote mouse pointer. This occurs because the resolution directly supported by JavaRConsole is 1024x768; the Windows display driver detects this, and simulates the larger display using panning. This confuses the local mouse driver.

Workaround

Simply ignore the local mouse pointer while interacting with the remote system. Alternately, you can use a display resolution of 1024x768 or less.

Special Character in FTP Password Disables Snapshot Feature (6674741)

The ILOM snapshot feature does not work if an FTP password with special characters is used. For example, suppose an FTP account on server “ftp” has a user name of “user”, a password of “abc@123”. You should be able to specify that account as a snapshot with the following command:

set dump_uri=ftp://user:abc%40123@ftp/

In this URI, the “%40” represents the “@” character. This syntax is not interpreted correctly, and the FTP login fails.

Service Processor Does Not Log Event or Provide Visual Alert After Hard Disk Is Removed

If a hard disk is removed from a system, the service processor (SP) neither logs an event nor provides a visual alert. This is expected behavior because the service processor does not receive events from the LSI SAS controller when a disk drive is removed and visual alerts do not occur.

Recovery For Accidentally Flashing Wrong Platform Firmware on Sun Fire X4600 M2 Server

If you accidentally flash the wrong platform firmware onto a Sun Fire X4600 M2 server, use IPMItool to verify the hardware version. For example, if the SP IP address is 10.0.0.13 and the root password is “changeme”, you might enter the command as follows (boldface indicates user input):

# ipmitool -H 10.0.0.13 -U root -P changeme fru
FRU Device Description : p0.card.fru (ID 10)
Board Product : ASSY,CPU BOARD,X4600
Board Serial : 002555
Board Part Number : 501-7321-03
Board Extra : 01
Board Extra : G4_CPU

If the Board Product line lists something other than ASSY,CPU BOARD,X4600 M2, then the server has the wrong firmware. For example, if it lists ASSY,CPU BOARD,X4600, then you have accidentally flashed the Sun Fire X4600 firmware onto the Sun Fire X4600 M2.

The system will not reboot, but the service processor is still functional; you can return to the Sun Fire X4600 M2 firmware.

1. Download and flash upgrade the server to the latest version of the Sun Fire X4600 M2 firmware.

2. Reboot the server.

3. Use IPMItool to verify the hardware version, as shown above.

# ipmitool -H 10.0.0.13 -U root -P changeme fru

If the Board Product line includes X4600 M2, then the server has the correct firmware.

Unimplemented Simple Network Management Protocol (SNMP) Traps (6300437)

The following traps are currently not supported:

• SUN-PLATFORM-MIB
• sunPlatObjectCreation
• sunPlatObjectDeletion
• sunPlatCommunicationsAlarm
• sunPlatEnvironmentalAlarm
• sunPlatEquipmentAlarm
• sunPlatProcessingErrorAlarm
• sunPlatStateChange
• sunPlatAttributeChangeInteger
• sunPlatAttributeChangeString
• sunPlatAttributeChangeOID
• sunPlatQualityOfServiceAlarm
• sunPlatIndeterminateAlarm

• ENTITY-MIB
• entConfigChange

• SNMPv2-MIB
• coldStart
• warmStart
• authenticationFailure

SNMP Agent Does Not Handle All SETs (6255301)

The SNMP agent does not currently handle SETs for the Entity and Sun Platform MIBs. This causes compliance tests involving SETs to fail. Use other services to perform the needed tasks.

---

Resolved Service Processor Issues

SP Does Not Automatically Notify User When SSL Certificate is Updated (6476473)

(Fixed in Software 1.1.)

To establish a secure HTTPS connection to the ILOM, you must upload an SSL certificate and a private key into the ILOM. When the certificate is uploaded through the web, it is put in /conf/actualcert.pem on the SP. Similarly, the private key is put in /conf/actualprivkey.pem. The SP must be reset for the update to be applied. See the Integrated Lights-Out Manager (ILOM) Administration Guide (819-1160) for more information on updating SSL certificates.

After the SP resets, a pop-up message should inform the user that the keys have changed and prompt whether to accept them for a one-time login, or permanently. However, this pop-up message is not displaying for Sun Fire X4600 M2 servers.

To view whether the SSL certificate and key were updated, use the following procedure:

1. Quit and restart your browser.

2. Use the browser to connect to the server’s IP address (using HTTPS).

3. Accept a pop-up message that indicates that the URL does not match the hostname in the certificate.

4. Click the View Certificate button to see the SSL certificate version that the server is using.

Web Interface Hangs After Change to Network Settings (6496326)

(Fixed in Software 1.2.)

After the Web interface is used to change the service processor network settings, the Web GUI might hang.

Workarounds

• Use the CLI or IPMItool to reconfigure the network interface.
• After using the Web GUI to reconfigure the network interface, restart the browser.

Service Processor Incorrectly Indicates Power Supply Fault After Shutdown (6516944)

(Fixed in Software 1.2.)

On systems running Service Processor firmware 1.1.1, the power supply fault LED is lit when the system is powered down, even if there is no power supply problem. The LED behaves correctly when the system is powered up.

Workarounds

• Power up the system and re-check the power supply fault LED.
• Query the IPMI Sensor Data Record (SDR). (The SDR is accessible through IPMI, the SP GUI, and the SP command line.) Refer to the psX.prsnt (present) and the psX.vinok (voltage in OK) indicators.
• Query the IPMI System Event Log (SEL) to determine if critical power supply or voltage events have occurred.

JavaRConsole Does Not Notify User Of SP Disconnect Right Away (6487885)

(Fixed in Software 1.3)

JavaRConsole continues to work (for example, redirecting a CD-ROM image) even after it is disconnected from the service processor. When the service processor is reset, the user is not informed that the connection has been lost until the service processor is finished rebooting.

Workaround

When a service processor is reset, or its network configuration changes, restart the client application or the browser running the Web GUI.

System Hangs During POST (6618895)

(Fixed in Software 1.3.)

The system might hang during POST. Often the system freezes shortly after the console displays the manufacturer’s logo. This problem is caused by a a bug in the ILOM software that supports configuring the Serial Management port. This problem can be triggered by referencing the /SP/serial part of the ILOM object tree from the ILOM command line. Avoid commands such as these:

show /SP/serial/host
show -level all

Also avoid using the Configuration/Serial Port tab in the ILOM Web interface.

Workaround

To work around this problem, clear the internal software state that causes it. Use any of the following methods:

• Display the Configure/Serial Port tab in the ILOM Web interface. There is no need to change any options; simply displaying the tab toggles the condition
• From the ILOM command line, enter this command:


show /SP/serial/external


• Reset the Service Processor.

Cannot Set Syslog IP Address When Zeros Are Used (6547470)

(Fixed in Software 1.3.)

An attempt to set the syslog IP address fails if the address contains an octet of 0. For example, 10.10.10.10 and 100.100.100.100 works, while 19.52.0.17 does not work.

Unable to Set Multiple Properties in a Single Command (6544145)

(Fixed in Software 1.3)

The ILOM CLI sometimes reports an internal error for a command that attempts to set multiple properties.

Workaround

Set properties one at a time.

CLI Cannot Set ntp Server Address to Some IP Addresses (6543859)

(Fixed in Software 1.3)

The ILOM CLI cannot set the ntp server address to an IP address with a 0 as the second or third octet, such as 10.13.0.20 or 10.0.60.20. These addresses are rejected as “invalid.”

Workaround

Use the Web interface to set the ntp server address.

Uploading SSL Certificate Without the Key Causes Port 443 to Not Open (6550394)

(Fixed in Software 1.3.)

If the SSL certificate is uploaded without the key, it causes 443 to not open.

Workaround

Errors uploading the SSL certificate can disable the https interface. Use the CLI or ipmi to reflash the ILOM SP. Do not preserve the configuration.

CLI Sessions Can Lose Data Due to TFTP Timeout (6544091)

(Fixed in Software 1.3.)

Workaround

Close the CLI session and open a new one.

Reported Fan Speed is Always Non-Zero, Even When Fan Has Failed or Been Removed (6493168)

(Fixed in Software 1.3.)

The ILOM SP always reports a non-zero fan speed, even when the fan has failed or been removed.

Workaround

Use the fan RPM level and the lower non-recover IPMI event in the IPMI sel log to determine that a fan has failed or been removed.

System Powers Off During Reboot After Thermal Shutdown (6494327)

(Fixed in Software 1.3.)

If overheating forces a system shutdown, the system might power off during reboot.

Workaround

After a thermal shutdown, be sure to give the system enough time to cool off (up to 30 minutes). If the system still powers down, power it back up. It might take 2 or 3 tries to reboot.

Web Interface Power Supply Alert Doesn’t Work (6494817)

(Fixed in Software 1.3.)

The Web interface does not update the PWROK and VINOK indicators when power is lost.

Workaround

Use IPMItool output to detect loss of power.

Web Interface Hangs After Changing SP’s IP Address (6496327)

(Fixed in Software 1.3.)

When the Web interface is used to changes the service processor’s static IP address, the Web GUI might hang.

Workaround

Reset the service processor to reestablish the connection. To avoid this problem, do not use the Web GUI to change the IP address; use IPMItool or the CLI instead.

Cannot Set ipgateway to 0.0.0.0 (6675084)

(Fixed in Software 2.1.)

The ipgateway target cannot be set to 0.0.0.0. Attempting to do so may cause existing network settings to be lost.

Web Interface May Be Temporarily Unavailable after Changing Network Settings (6669412)

(Fixed in Software 2.1.)

Changing the network setting can cause the Web interface to become temporarily unavailable for up to 20 minutes. The ILOM’s web server process blocks, and web client requests are not honored.

Workaround

Kill and restart the Web interface process.

Cannot Start Services (6544951)

(Fixed in Software 2.2.)

The ILOM may be unable to start services when the maximum number of users or communities are configured, or multiple commands are issued in a short period.

Workaround

Do not configure the maximum number of users or communities. Wait several seconds between configuration commands.

Control Keys to Host Console via ILOM CLI During POST Scan of Onboard NIC Can Hang System (6767733)

(Fixed in Software 2.3.)

The host console can be viewed remotely using the following ILOM CLI command:

start /SP/console

If the host console is opened this way while the system is being powered up, pressing Control-S, Control-E, Control-P, or Control-N during the POST scan of the onboard NIC can cause the system to hang.

---

BIOS Issues

The following are current issues.

Upgrade Past Software 2.2 Causes Device Enumeration For PCI Slots 5, 6, and 7 to Change (6719935)

Cannot Boot From Internal Disk With Sun Multithreaded 10 GbE Card in Slot 2, 3, or 4 (6711405, 6764001)

Erroneous Report that PowerNow! Not Supported (6768326)

Abnormal and Intermittent Delays on Boot on 6P and 8P Systems (6734988)

Option ROM Space For PXE Booting Can Be Exhausted Before All Devices Can Be Scanned (6453144, 6403173, 6272514, 6393809, 6439856, 6462303)

Twelve-Boot-Device Limitation Can Be Exceeded Before All Devices Can Be Scanned By BIOS (6268877, 6439856)

Upgrade Past Software 2.2 Causes Device Enumeration For PCI Slots 5, 6, and 7 to Change (6719935)

A change in the system BIOS, introduced after Software 1.3, changes the way the system enumerates PCIe cards in slots 5, 6, and 7, as well as the devices connected to these cards. Upgrading the BIOS can cause the names of disk partitions and network interfaces to change.

Workaround

BIOS 126, included with Software 2.3, contains a setting that restores the old device enumeration. After flashing new system firmware, but before starting the host operating system, enable the following system BIOS option:

• Advanced/Hardware Error Handling/Unused PCIE Root Port

Cannot Boot From Internal Disk With Sun Multithreaded 10 GbE Card in Slot 2, 3, or 4 (6711405, 6764001)

A system with a Sun Multithreaded 10Gb Ethernet card (X1027A-z or X1107A-z) in slot 2, 3, or 4 is unable to boot from the local disk. This problem occurs because the Ethernet card has a large option ROM. When the Ethernet card is scanned first before the embedded HBA, there is not enough space left to load the HBA option ROM, and the internal disks become inaccessible.

Workaround

There are two workarounds:

• Install the Ethernet card in a slot other than 2, 3, or 4.
• Disable any unneeded option ROM.

Erroneous Report that PowerNow! Not Supported (6768326)

The Red Hat Enterprise Linux 5.2 XEN kernel reports the following in the system event log:

powernow-k8: Your BIOS does not provide _PSS objects.  PowerNow! does not work on SMP systems without _PSS objects.  Complain to your BIOS vendor.

This message is not correct and can be safely ignored.

Abnormal and Intermittent Delays on Boot on 6P and 8P Systems (6734988)

A system with 6 or 8 CPU modules normally experiences a delay of 1 to 2 minutes between power-on or reset and the appearance of the first BIOS messages on the console display. Some systems experience intermittent delays beyond 2 minutes. These are the symptoms:

• Delay occurs before any BIOS message appears on the system console.
• The last message in the IPMI System Event Log is “System Firmware Progress | Primary CPU initialization | Asserted”. If the subsequent message “System Firmware Progress | Keyboard controller initialization | Asserted” appears, the delay is over or did not occur.
• No system delays consistently. A system will rarely delay for a long period twice during two successive boots.
• The length of the delay varies considerably. The system may resume booting after a few minutes, or it may not resume for hours.

Workaround

If the console remains blank for 4 minutes or more, reset the system. In most cases it will take one reset to get past the delay; in a few cases it will take two.

Option ROM Space For PXE Booting Can Be Exhausted Before All Devices Can Be Scanned (6453144, 6403173, 6272514, 6393809, 6439856, 6462303)

The amount of BIOS option ROM space is limited and can be exhausted when many option cards are installed in the system.

This is expected behavior.

Possible Effects:

• Cannot PXE boot through the Ethernet ports and the following error message is displayed: Not enough space to copy PCI Option ROM
• Cannot PXE boot through the Ethernet ports and the following error message is displayed: Base-Code ROM ID structure not found
• PCI cards cannot PXE boot as desired if they are installed in a PCI slot that is scanned after the option ROM space is exhausted.
• HBA card and its attached HDD array are not detected if installed in a PCI slot that is scanned after the option ROM space is exhausted.

The devices and PCI slots are detected by the BIOS during startup in the following order (also see FIGURE 2-2 for the placement of the PCI slots):

1. PCI-E slot 2

2. PCI-E slot 3

3. PCI-E slot 4

4. On-board Intel NIC

5. PCI-X slot 0

6. PCI-X slot 1

7. On-board LSI SCSI controller

8. PCI-E slot 5

9. PCI-E slot 6

10. PCI-E slot 7

Workarounds

There are two possible workarounds to ensure that you have enough option ROM space to PXE boot from your devices as desired.

• If the device you want to boot from appears in the list of boot devices in the BIOS, perform Option 2 below to change the scanning order.
• If the device you want to boot from does not appear in the list of boot devices, perform Option 1 below so that the device appears in the list, then perform Option 2 to change the scanning order.

Disable option ROM scanning on all devices that do not need to PXE boot. This will preserve the option ROM space for the devices that you do want to PXE boot. Use the following procedure.

1. Enter the BIOS Setup utility by pressing the F2 key while the system is booting up and performing POST.

2. On the BIOS Main Menu screen, select the PCIPnP tab to open the PCI/PnP Settings screen.

3. Change the fields to Disabled for those PCI cards or NICs that will not be PXE booted.

4. Press and release the right arrow key until the Exit menu screen is displayed.

5. Follow the instructions on the Exit menu screen to save your changes and exit the Setup utility.

Manually set the BIOS boot order so that the devices that you want to PXE boot from are early enough in the boot order to be scanned before the option ROM space is exhausted. Use the following procedure:

1. Enter the BIOS Setup utility by pressing the F2 key while the system is booting and performing POST.

2. On the BIOS Main Menu screen, select the Boot tab to open the Boot menu main screen.

3. Select Boot Device Priority, or select Hard Disk Drives from the list to change hard-disk drives.

4. Change the selections for the boot devices or hard-disks drives to set the required device order.

5. Press and release the right arrow key until the Exit menu screen is displayed.

6. Follow the instructions on the Exit menu screen to save your changes and exit the Setup utility.

FIGURE 2-1 PCI Designation and Speeds of PCI Slots

[ D ]

Twelve-Boot-Device Limitation Can Be Exceeded Before All Devices Can Be Scanned By BIOS (6268877, 6439856)

The server BIOS can scan up to 12 devices, but depending on what is installed in the server, this limit can be exceeded before all devices in the server can be scanned. For example, if you install an HBA with an attached disk array into a PCI slot that is scanned before the on-board LSI SAS controller, the internal HDDs might not be scanned if the HBA and its attached HDDs exceed the 12-device limit.

This is expected behavior. Devices in excess of the 12-device limit will be seen at the OS level, but are not scanned by the BIOS during POST.

Possible Effects:

• Cannot boot from internal HDDs.
• NICs are not scanned and therefore cannot be used for PXE booting.

The BIOS scans devices in ascending order (from low PCI address to high PCI address). The scanning priority is:

1. PCI-E slot 2

2. PCI-E slot 3

3. PCI-E slot 4

4. On-board Intel NIC

5. PCI-X slot 0

6. PCI-X slot 1

7. On-board LSI SCSI controller

8. PCI-E slot 5

9. PCI-E slot 6

10. PCI-E slot 7

Workaround

Perform Option 1 below so that the device appears in the list, then perform Option 2 to change the scanning order.

Disable option ROM scanning on all devices that do not need to PXE boot. This will preserve the option ROM space for the devices that you do want to PXE boot. Use the following procedure.

Enter the BIOS Setup utility by pressing the F2 key while the system is booting up and performing POST.

1. On the BIOS Main Menu screen, select the PCIPnP tab to open the PCI/PnP Settings screen.

2. Change the fields to Disabled for those PCI cards or NICs that will not be PXE booted.

3. Press and release the right arrow key until the Exit menu screen is displayed.

4. Follow the instructions on the Exit menu screen to save your changes and exit the Setup utility.

Manually set the BIOS boot order so that the devices that you want to PXE boot from are early enough in the boot order to be scanned before the option ROM space is exhausted. Use the following procedure:

1. Enter the BIOS Setup utility by pressing the F2 key while the system is booting and performing POST.

2. On the BIOS Main Menu screen, select the Boot tab to open the Boot menu main screen.

3. Select Boot Device Priority, or select Hard Disk Drives from the list to change hard-disk drives.

4. Change the selections for the boot devices or hard-disks drives to set the required device order.

5. Press and release the right arrow key until the Exit menu screen is displayed.

6. Follow the instructions on the Exit menu screen to save your changes and exit the Setup utility.

---

Resolved BIOS Issues

Quad-Core System Won’t Boot When Hypertransport Coherent Link Widths Are Set To 2 Bits (6710259)

IO Address Conflict (6540138)

Uncorrected ECC Error Reported on Wrong DIMM Pair on 8-DIMM CPU Module (6509975)

Random Panics Might Occur When HPET Timer is Enabled (6478355)

Quad-Core System Won’t Boot When Hypertransport Coherent Link Widths Are Set To 2 Bits (6710259)

(Fixed in Software 2.3)

If the BIOS settings for Hypertransport Coherent Link Width are 2 bits, then a system with quad-core processors

Workaround

In BIOS setup, set each of the Hypertransport Coherent Link Widths to a value other than “2” or “4”.

IO Address Conflict (6540138)

(Fixed in Software 2.0)

IO address conflicts might occur in systems with 2 PCI-X and 6 PCI-E slots populated, with a Sun PCI-E Quad Gigabit Ethernet Card (X4446A-Z) in slot 5, 6, or 7. Systems running Solaris OS might reset before the boot sequence is complete, leading to a reboot cycle.

Workaround

Move the Quad Gigabit Ethernet Card to slot 2, 3, or 4.

Uncorrected ECC Error Reported on Wrong DIMM Pair on 8-DIMM CPU Module (6509975)

(Fixed in Software 2.0)

When an uncorrected ECC error occurs on an 8-DIMM CPU module, the error might be reported on the lowest-numbered DIMM pair (the pair closest to the CPU) regardless of where the error actually occurred. For example, in a fully-populated module, an error in Pair 3 (DIMM7 and DIMM6), Pair 2 (DIMM5 and DIMM4), or Pair 1 (DIMM3 and DIMM2) may be reported as occurring in Pair 0 (DIMM1 and DIMM0).

This problem can cause the system to reboot repeatedly, with no output on the video display. If this occurs, check the ILOM System Event Log for frequent ECC errors in a lowest-numbered pair.

Workarounds

• Carefully reseat the CPU. Check for bent or dirty pins and for misplaced thermal grease.
• Replace one or more higher-numbered pairs with known-good DIMMs.
• Carefully reseat all DIMMs on the module.

Random Panics Might Occur When HPET Timer is Enabled (6478355)

(Fixed in Software 1.2)

The interaction of some operating systems with the high-performance event timer (HPET) on this server might result in system panics (reboots). When this happens, the following message appears after the last POST window during a boot process, and in the /var/log/messages file:

MT-BIOS bug: 8254 timer not connected to IO-APIC

If this problem occurs, you can disable the HPET timer. This setting is in the BIOS Configuration Utility’s Advanced menu, on the ACPI Configuration Screen.

1. Enter the BIOS Setup utility by pressing the F2 key while the system is booting up and performing POST.

2. On the BIOS Main Menu screen, select the Advanced tab to open the Advanced menu screen.

3. On the Advanced menu screen, select ACPI Configuration.

4. On the ACPI configuration screen, select Advanced ACPI Configuration.

5. Change the field to Disabled for ACPI HPET Support.

6. Press and release the right arrow key until the Exit menu screen is displayed.

7. Follow the instructions on the Exit menu screen to save your changes and exit the Setup utility.

---

LSI Firmware Issues

The following are current issues.

Error Message After Disk Swap (6766054)

In testing, error messages appeared for disk swaps on a RAID1 or RAID1E array created using the onboard LSI controller BIOS:

This message does not reflect a real RAID problem and can be safely ignored.

RAID Volume Requires 64 Mbytes of Unpartitioned Hard-Disk Space for Metadata (6312581)

To create a RAID volume, the firmware and BIOS must write metadata at the end of the hard-disk drive (HDD). At least 64 Mbytes of unpartitioned hard-disk space are required.

In servers shipped with the preinstalled Solaris 10 1/06 (and later) OS, the preinstall image includes more than 64 Mbytes of unpartitioned hard disk space for metadata. No further action is required.

Sun Fire X4600 M2 Server Product Notes

819-5032-24

Copyright © 2010, Oracle and/or its affiliates. All rights reserved.