Netra SPARC T3-1B Server Module

Late-Breaking Information

This chapter provides the following late-breaking information for the Netra SPARC T3-1B server module:

• Hardware Issues
• System Firmware Issues
• Oracle ILOM Issues
• Oracle Solaris Issues
• Related Documentation

---

Hardware Issues

Time Changes After System Reset (CR7127740)

After a cold reset, the server might add one day to the Oracle Solaris OS date and time. This date change will only occur on the first cold reset after January 1, 2012.

Workaround: Obtain firmware version 8.1.4.e or newer.

Server Module Might Panic During Hot-Plugging of the 10GbE Dual Port SFP+ ExpressModule
(CR 6974235)

Initiating hot-plug removal of a 10GbE Dual Port SFP+ ExpressModule connected to the server module by using the ExpressModule ATTN button might cause the Oracle Solaris OS that is running on the server module to panic.

You can safely use the ATTN button to hot-plug insert the same Express Module into a slot that connects to the server module.

Workaround: Do not use the PCIe EM ATTN button for hot plug removal. Instead, use the Oracle Solaris cfgadm(1M)command to hot-plug remove a 10GbE Dual Port SFP+ Express Module connected to the server module.

For example, to hot-plug remove an express module in slot PCI-EM1 using cfgadm, perform the following commands in the Oracle Solaris instance connected to the ExpressModule:

# cfgadm -c unconfigure PCI-EM1
# cfgadm -c disconnect  PCI-EM1

Replace Faulty DIMMs With Uncorrectable Errors as Soon as Possible (CR 6996144)

This issue is fixed in the Oracle Solaris 10 8/11 OS.

If a DIMM has an uncorrectable error (UE), the server will generate a fault.memory.bank error that labels a DIMM as faulty. You can view this error using the Oracle ILOM show faulty command or using the fmdump -v command.

If a DIMM contains a persistent UE (an error that continually occurs even after multiple reboots), replace this DIMM as soon as possible to avoid any server downtime.

Workaround: Replace faulty DIMMs as soon as possible.

Not Oracle Certified DIMM Warning Message (CR 7034912)

After installing supported optional component DIMMs shipped from Oracle Corporation or from a certified Oracle reseller, or after replacing a failed DIMM with a FRU DIMM, you might see warning messages similar to the following:

[CPU 0:0:0:] WARNING: /SYS/MB/CMP0/BOB0//CH0/D0: Not Oracle Certified

The system displays these messages because optional component and FRU DIMMs have not been marked as certified. Oracle certifies only DIMMs that ship installed in a system from the factory. Although Oracle has not certified these DIMMs, they are still supported. You can safely ignore these warning messages.

---

System Firmware Issues

Link Width x8 Link Speed GEN1 Warning Displayed During Power Up (CR 6958263)

This is fixed in SysFW 8.0.4.c and higher.

On rare occasions while powering up the server module, the following error message might be displayed just before the system reaches the OBP prompt:

WARNING: ios0, peun Link Width x8 Link Speed GEN1.

Workaround: Reset the system at the OBP prompt.

ok reset-all

---

Oracle ILOM Issues

picld Messages Logged (CR 6992903)

Occasionally, these warning messages might be logged in the /var/adm/messages file:

picld[177]: [ID 629468 daemon.warning] PICL snmpplugin: cannot get entPhysicalName (row=xxx)

Workaround: These messages are harmless, and you can safely ignore them.

---

Oracle Solaris Issues

prtdiag Labels the Server Module Serial Number as the Chassis Serial Number
(CR 6669159)

The last few lines of the prtdiag -v output display the server module serial number as the Chassis Serial Number. This label is misleading because it might be interpreted as the modular system chassis serial number.

Workaround: Be aware that the prtdiag Chassis Serial Number is the server module serial number.

Oracle VTS disktest Might Fail on USB Devices (CR 6873719)

USB drives connected to the front dongle cable or the internal USB port might fail after running the Oracle VTS disktest for over 5 hours.

Workaround: Stop any application that is using the USB port. Then reset the USB port with the cfgadm -x usb_reset [...] command.

False nxge Warning Messages (CR 6938085)

This issue is fixed in the Oracle Solaris 10 8/11 OS.

During the normal operation of your server, you might see the following warning messages in the system console or in the Oracle Solaris /var/adm/messages file:

date/time machinename nxge: [ID 752849 kern.warning] WARNING: nxge0 : nxge_hio_init: hypervisor services version 2.0

Workaround: These messages are harmless, and you can ignore them.

fault.memory.memlink-uc Interconnect Fault Did Not Cause Panic as Stated by Knowledge Article (CR 6940599)

When a fault.memory.memlink-uc interconnect fault is detected, the system should shut down to protect memory integrity. On intermittent occasions, this fault has been reported during boot operations without the system shutting down.

Although it is possible that this irregular behavior indicates that the system was able to recover from the memory link error and restore a healthy boot-up state, the safest action is to perform a power-down and power-up sequence.

Recovery: Power cycle the server module.

Degraded Network Performance When Using Sun Dual 10GbE PCIe EMs (CR 6943558)

Excessive packet loss can occur when two or more ports are used across multiple Sun Dual 10GbE PCIe 2.0 PCIe EMs. As a result, transmit and receive performance is significantly degraded.

Workaround: Enable flow control in the ixgbe driver by performing the following procedure. This action greatly reduces packet loss and improves performance.

As superuser, add the following line in the /kernel/drv/ixgbe.conf file:

flow_control = 3;

Then reboot the server module.

unsupported port mode Messages Logged
(CR 6962912)

When a 10GbE XAUI Pass-Through FEM is used with an incompatible NEM (such as a NEM and FEM combination not listed in ‘Supported Modular Components), the following messages might be logged in the /var/adm/messages file:

nxge: [ID 752849 kern.warning] WARNING: nxge1 : nxge_n2_kt_serdes_init:port<0> - unsupported port mode 8
nxge: [ID 752849 kern.warning] WARNING: nxge1 : nxge_n2_kt_serdes_init:port<1> - unsupported port mode 8
nxge: [ID 752849 kern.warning] WARNING: nxge1 : nxge_n2_serdes_init: Failed to initialize N2 serdes for port<0>
nxge: [ID 752849 kern.warning] WARNING: nxge1 : nxge_n2_serdes_init: Failed to initialize N2 serdes for port<1>

Workaround: Ensure that your server module is configured with the correct NEM and FEM combination. These messages are not an indication of malfunctioning hardware or software. You can ignore the messages.

Spurious Interrupt Message in System Console (CR 6963563)

During normal operation and when running the Oracle VTS system exerciser, you might see the following message in the system console or in the /var/adm/messages file:

date time hostname px: [ID 781074 kern.warning] WARNING: px0: spurious interrupt from ino 0xn
date time hostname px: [ID 548919 kern.info] ehci-0#0
date time hostname px: [ID 100033 kern.info]

Workaround: You can safely ignore these messages.

Spurious Error Message During Initial Oracle Solaris OS Installation (CR 6971896)

This issue only occurs when you are performing an installation using a keyboard, mouse, and monitor.

The miniroot is a bootable root file system that includes the minimum Oracle Solaris OS software required to boot the server module and configure the OS. The miniroot runs only during the installation process. When the server module boots the miniroot for the initial configuration, you might see the following messages in the system console:

Fatal server error:
InitOutput: Error loading module for /dev/fb
giving up.
/usr/openwin/bin/xinit: Network is unreachable (errno 128):
unable to connect to X server
/usr/openwin/bin/xinit: No such process (errno 3): Server error.

The messages indicate that the Xsun server in the Oracle Solaris OS miniroot cannot find a supported driver for the AST graphics device in the service processor.

These messages are expected because the miniroot contains only the Xsun environment, and the AST frame buffer (astfb) is supported only in the Xorg environment. The Xorg environment is included in the installed OS. Therefore, you can use the graphics device when running the installed OS.

Workaround: You can safely ignore these messages.

Hot-Plug Removal of PCIe EMs Might Generate devfsadmd Errors (CR 6973637)

For PCIe EMs connected to this server module, using the ATTN button to prepare a PCIe EM for hot-plug removal might generate the following error:

devfsadmd[202]: failed to lookup dev name for /pci@400/pci@2/pci@0/pci@1/.....

Workaround: You can safely ignore these errors.

Error Report Event (ereport) Not Generated for a Degraded Service Processor (CR 6978171)

This issue is fixed in the Oracle Solaris 10 8/11 OS.

The following incorrect event error report (ereport) is generated if the service processor is operating in a degraded state:

ereport.fm.fmd.module

However, a degraded service processor should generate the following ereport:

ereport.chassis.sp.unavailable

To view ereport events, use the fmdump -eV command. Refer to the fmdump(1M) man page for instructions.

Workaround: Clear all service processor faults to ensure that the service processor operates in a normal state.

False Error Report Events (ereport) Generated When System Is Booted With a Degraded Service Processor (CR 6981297)

This issue is fixed in the Oracle Solaris 10 8/11 OS.

If the system boots with a degraded service processor, the system generates an error report event (ereport) that does not accurately state the problem. The ereport should state that it cannot make a connection to the service processor instead of the following false ereport:

msg = error: bad conn open during ver negot: errno 5

To view ereport events, use the fmdump -eV command. Refer to the fmdump(1M) man page for instructions.

Workaround: Clear all service processor faults to ensure that the service processor operates in a normal state.

Oracle Enterprise Manager Process Hangs and Becomes Unkillable (CR 6994300)

This issue is fixed in the Oracle Solaris 10 8/11 OS.

The Oracle Enterprise Manager Java process can hang and become unkillable on the server module. When the Enterprise Manager process hangs, it continues to listen on its web UI port, which makes the process unkillable. This problem has been seen on servers running both the Java SE 5.0 version that is bundled with Oracle database software and the most recent downloadable Java SE 6 Update 22 version.

Workaround: Reboot the system. If the problem repeats, contact your authorized service provider.

nxge Driver Not Loaded (CR 6995458)

If Oracle Solaris 10 10/09 is installed on the server module along with Oracle Solaris 10 9/10 Patch Bundle, the SPARC T3 NIU ports might be unusable on Oracle Solaris.

Workaround: Add this entry to the /etc/driver_aliases file and then reboot Oracle Solaris:

nxge "SUNW,niusl-kt"

---

Related Documentation

Netra SPARC T3-1B Server Module

E21654-04

Copyright © 2012, Oracle and/or its affiliates. All rights reserved.