Sun SPARC® Enterprise M8000/M9000 Servers Product Notes For XCP Version 1070 Sun Microsystems, Inc. www.sun.com Part No. 820-4293-10 April 2008, Revision A Submit comments about this document at: http://www.sun.
Copyright 2008 Sun Microsystems, Inc., 4150 Network Circle, Santa Clara, California 95054, U.S.A. and FUJITSU LIMITED, 1-1, Kamikodanaka 4-chome, Nakahara-ku, Kawasaki-shi, Kanagawa-ken 211-8588, Japan. All rights reserved. Sun Microsystems, Inc.
Copyright 2008 Sun Microsystems, Inc., 4150 Network Circle, Santa Clara, California 95054, U.S.A. et FUJITSU LIMITED, 1-1, Kamikodanaka 4-chome, Nakahara-ku, Kawasaki-shi, Kanagawa-ken 211-8588, Japon. Tous droits réservés. Entrée et revue tecnical fournies par Fujitsu Limited sur des parties de ce matériel. Sun Microsystems, Inc. et Fujitsu Limited détiennent et contrôlent toutes deux des droits de propriété intellectuelle relatifs aux produits et technologies décrits dans ce document.
Contents Preface vii Sun SPARC Enterprise M8000/M9000 Servers Product Notes 1 New in XCP 1070 1 Supported Firmware and Software Versions Using a WAN Boot Server Solaris Patch Information 2 3 Installing the Solaris Patches Upgrading to XCP 1070 2 3 4 General Functionality Issues and Limitations Limitations for SPARC64 VII Processors 4 4 General Functionality Issues and Limitations Hardware Installation and Service Issues Software and Firmware Issues 5 6 6 XCP Issues and Workarounds 6 Sola
Sun Management Center Software Issues and Workarounds Software Documentation Updates 17 Upgrading From XCP 1041 or Lower 19 ▼ To Prepare to Upgrade ▼ To Upgrade From XCP 1041 or Lower Additional Software Procedures 19 27 To Upgrade the wanboot Executable Identifying Degraded Memory in a System ▼ 21 27 Booting From a WAN Boot Server ▼ 27 28 To Identify Degraded Memory in a System 28 Identifying Different Memory Sizes in a System Board 28 ▼ To Use the showdevices Command ▼ To Use the p
Preface These product notes contain important and late-breaking information about the Sun SPARC® Enterprise M8000/M9000 servers hardware, software, and documentation. Technical Support If you have technical questions or issues that are not addressed in the Sun SPARC Enterprise M8000/M9000 servers documentation, contact your local Sun™ Service representative. For customers in the U.S. or Canada, call 1-800-USA-4SUN (1-800-872-4786).
Sun Java Enterprise Server The Sun Java Enterprise Server is a comprehensive set of software and lifecycle services that make the most of your software investment. For an overview and documentation, go to: http://www.sun.com/service/javaes/index.xml Note – Due to an issue that arises from the installation of the Java Enterprise System 5 Update 1 on your system (CR 6644798), it might be necessary to enable the WebConsole SMF service.
The Sun Connection Update Manager can be used to reinstall the patches if necessary or to update the system with the latest set of mandatory patches. For more Information about the Sun Connection Update Manager, refer to the Sun Update Connection System Administration Guide at: http://docs.sun.com/app/docs/prod/updconn.sys Or visit: http://wikis.sun.
Note – Patch 118833-xx is a kernel patch that requires special instructions for installation (see the patch README for specifics) and therefore is a download-only (interactive) patch requiring manual installation. You must install patch 118833-xx first in order for any remaining patches in the patch set to be installed. 5. For a kernel patch such as 118833-xx, ontinue by typing: # cd /var/sadm/spool # unzip patchid-xx.jar 6.
2. Edit the file /tmp/RegistrationProfile.properties to add your user name, password, network proxy (if necessary), and port (if required). Note – The user name and password is a Sun Online Account. To create an account, go to http://sunsolve.sun.com. 3. Register your system by typing: # sconadm register -a -r /tmp/RegistrationProfile.properties 4. Obtain the correct patches for your system by typing: # smpatch set patchpro.patchset=sem4k5k8k9k 5.
8. Download and install the patches by typing: # smpatch update 9. If any of the patches requires a system restart, see Step 6. The patch installation is now complete. Additional Information For additional information, see the release notes for the version of the Solaris OS that you are using, as well as the Big Admin web site: http://www.bigadmin.
Sun Welcomes Your Comments Sun is interested in improving its documentation and welcomes your comments and suggestions. You can submit your comments by going to: http://www.sun.
xiv SPARC Enterprise M8000/M9000 Servers Product Notes for XCP 1070 • April 2008
Sun SPARC Enterprise M8000/M9000 Servers Product Notes This document includes these sections: ■ “New in XCP 1070” on page 1 ■ “Supported Firmware and Software Versions” on page 2 ■ “Solaris Patch Information” on page 3 ■ “Upgrading to XCP 1070” on page 4 ■ “General Functionality Issues and Limitations” on page 4 ■ “Hardware Installation and Service Issues” on page 6 ■ “Software and Firmware Issues” on page 6 ■ “Software Documentation Updates” on page 17 ■ “Upgrading From XCP 1041 or Lower”
Supported Firmware and Software Versions TABLE 1 lists the minimum required versions of some supported software and firmware for XCP 1070 on Sun SPARC® Enterprise M8000/M9000 servers.
Solaris Patch Information Currently, patches are required only for servers running Solaris 10 11/06 OS. The following patches are required: ■ 118833-36 ■ 125100-10 ■ 123839-07 ■ 120068-03 ■ 125424-01 ■ 118918-24 ■ 120222-21 ■ 125127-01 ■ 125670-02 ■ 125166-05 These patch identifiers represent the minimum level of the patches that must be installed. The two-digit suffix represents the minimum revision level of the patch. Check SunSolve.Sun.
8. 125127-01 – Reboot your domain before proceeding. 9. 125670-02 10. 125166-05 Upgrading to XCP 1070 If you are upgrading to XCP 1070 from a version of XCP 1041 or lower, refer to “Upgrading From XCP 1041 or Lower” on page 19 for important instructions. If you are upgrading from a more recent version of XCP, refer to the Sun SPARC Enterprise M4000/M5000/M8000/M9000 Servers XSCF User’s Guide for instructions.
General Functionality Issues and Limitations Caution – For dynamic reconfiguration (DR) and hot-plug issues, see “Solaris OS Issues and Workarounds” on page 8. Note – For power-on after power-off, wait at least 30 seconds before turning the system power back on, by using the main line switch or the circuit breakers on the distribution panel. ■ DR and XSCF failover are not compatible. Do not start an XSCF failover while a DR operation is running.
■ Do not use the Service Processor (SP) as the Network Time Protocol (NTP) server. Using an independent NTP server provides optimal reliability in maintaining consistent time on the SP and the domains. For more information about NTP, see the Sun Blueprint document, Using NTP to Control and Synchronize System Clocks: http://www.sun.com/blueprints/0701/NTP.pdf Hardware Installation and Service Issues TABLE 3 lists known issues for which a defect change request ID has been assigned.
TABLE 4 XCP Issues and Workarounds ID Description Workaround 6565422 The Latest communication field in showarchiving is not updated regularly. Disabling and re-enabling archiving refreshes the Latest communication field in showarchiving output. 6575425 Most XSCF commands should display “Permission denied” when they are executed on the Standby XSCF. Instead, some commands report various errors.
Solaris OS Issues and Workarounds This section contains information about Solaris OS issues. TABLE 5, TABLE 6, and TABLE 7 list issues you might encounter, depending upon which Solaris OS release you are using. Solaris Issues for All Supported Releases TABLE 5 lists Solaris OS issues that you might encounter in any supported release of Solaris OS.
TABLE 5 Solaris OS Issues and Workarounds for All Supported Releases (2 of 4) CR ID Description Workaround 6531036 The error message network initialization failed appears repeatedly after a boot net installation. There is no workaround.
TABLE 5 Solaris OS Issues and Workarounds for All Supported Releases (3 of 4) CR ID Description 6589833 The DR addboard command might cause a There is no workaround. system hang if you are adding a Sun StorageTek Enterprise Class 4Gb Dual-Port Fibre Channel PCI-E HBA card (SG-XPCIE2FCQF4) at the same time that an SAP process is attempting to access storage devices attached to this card.
TABLE 5 Solaris OS Issues and Workarounds for All Supported Releases (4 of 4) CR ID Description Workaround 6625734 Systems with large number of processors in a single domain environment might have suboptimal performance with certain workloads. Use processor sets to bind application processes or LWPs to groups of processors. Refer to the psrset(1M) man page for more information. 6632549 fmd service on domain might fail to maintenance mode after DR operations.
Solaris Issues Fixed in Solaris 10 5/08 TABLE 6 lists issues that have been fixed in Solaris 10 5/08 OS. You might encounter them in supported releases earlier than Solaris 10 5/08. TABLE 6 Solaris OS Issues and Workarounds Fixed in Solaris 10 5/08 (1 of 4) CR ID Description Workaround 5076574 A PCIe error can lead to an invalid fault diagnosis on a large M9000/M8000 domain. Create a file /etc/fm/fmd/fmd.conf containing the following lines; setprop client.buflim 40m setprop client.
TABLE 6 Solaris OS Issues and Workarounds Fixed in Solaris 10 5/08 (2 of 4) CR ID Description Workaround 6545143 When kcage daemon is expanding the kcage area, if the user stack exists in the expanded area, its area is demapped and might cause a ptl_1 panic during the flushw handler execution. There is no workaround. 6545685 If the system has detected Correctable MemoryErrors (CE) at power-on self-test (POST), the domains might incorrectly degrade 4 or 8 DIMMs.
TABLE 6 Solaris OS Issues and Workarounds Fixed in Solaris 10 5/08 (3 of 4) CR ID Description 6559504 Messages of the form nxge: NOTICE: These messages can be safely ignored.
TABLE 6 Solaris OS Issues and Workarounds Fixed in Solaris 10 5/08 (4 of 4) CR ID Description Workaround 6584984 The busstat(1M) command with -w option might cause domains to reboot. There is no workaround. Do not use busstat(1M) command with -w option on pcmu_p.
TABLE 7 Solaris OS Issues and Workarounds Fixed in Solaris 10 8/07 (2 of 2) CR ID Description Workaround 6510861 When using the PCIe Dual-Port Ultra320 SCSI controller card (SG-(X)PCIE2SCSIU320Z), a PCIe correctable error causes a Solaris panic. Add the following entry to /etc/system to prevent the problem: set pcie:pcie_aer_ce_mask = 0x31c1 6520990 When a domain reboots, SCF might not be able to service other domains that share the same physical board.
Sun Management Center Software Issues and Workarounds TABLE 8 lists issues and possible workarounds for Sun Management Center software. TABLE 8 Sun Management Center Issues and Workarounds CR ID Description Workaround 6654948 When viewing the PlatAdmin System There is no workaround. Components table, you might experience a delay of about 26 minutes before an alarm is displayed. There is no actual error, just a delay.
TABLE 9 Software Documentation Updates (2 of 3) Document Page Number Change Sun SPARC Enterprise M4000/M5000/M8000/M9000 Servers XSCF User’s Guide Page 2-2 Section 2.1.1, “Setup Summary by the XSCF Shell.” Add the following note: Note - In addition to the standard default login, Sun SPARC Enterprise M4000/M5000/M8000/M9000 servers are delivered with a temporary login called admin to enable remote initial login, through a serial port. Its privileges are fixed to useradm and cannot be changed.
TABLE 9 Software Documentation Updates (3 of 3) Document Page Number Change Sun SPARC Enterprise M4000/M5000/M8000/M9000 Servers Administration Guide Page 70 “About Auditing” section. Add the following note at the end of the “Audit File Tools” section: Note - This chapter describes how to set up archived log files. The SP Security (SUNWspec) Package gives administrators and service providers a means to view those files.
20 DNS domain name nameserver :sun.com :100.200.300.400 interface status IP address netmask route :xscf#0-lan#0 :up :100.200.300.77 :255.255.254.0 :-n 0.0.0.0 -m 0.0.0.0 -g 100.200.300.1 interface status IP address netmask route :xscf#0-lan#1 :down : : : interface status IP address netmask :xscf#0-if :down : : interface status IP address netmask route route :lan#0 :down : : :-n 0.0.0.0 -m 0.0.0.0 -g 100.200.300. :-n 0.0.0.0 -m 0.0.0.0 -g 100.200.300.
The XSCF will be reset. Continue? [y|n] :n XSCF> setroute -c del -n 0.0.0.0 -m 0.0.0.0 -g 100.200.300.2 lan#0 XSCF> setroute -c del -n 0.0.0.0 -m 0.0.0.0 -g 100.200.300.1 lan#0 XSCF> applynetwork -y 2. Configure the ISN network. XCP 1050 or later supports dual XSCF configuration. The Inter-SCF Network provides an internal communication link between the two XSCF Units (active and standby).
1. Log in to the XSCF#0 using an account with platform administrative privileges. 2. Verify that there are no faulted or deconfigured components by using the showstatus(8) command. XSCF> showstatus No failures found in System Initialization. If any failures are listed, contact your authorized service representative before proceeding. 3. Power off all domains. XSCF> poweroff -a 4. Confirm that all domains are stopped: XSCF> showlogs power 5.
Caution – The flashupdate command will update one bank, reset the XSCF, and commence update of the second bank. Before proceeding to Step 9, you must verify that the current and reserve banks are both updated. If both banks indicate XCP revision 1070, proceed to the next step. 9. Confirm completion of the update. XSCF> showlogs event Confirm no abnormality happens while updating XCSF_B#0. 10. Confirm that both the current and reserve banks of XSCFU#0 display the updated XCP versions.
XSCF> rebootxscf The XSCF will be reset. Continue? [y|n] :y d. Wait until XSCF firmware reaches the ready state. This can be confirmed when the READY LED of the XSCF remains lit, or the following message appears on the serial console: XSCF Initialize complete 12. Turn off all the server power switches for 30 seconds. 13. After 30 seconds, turn the power switches back on. 14. Wait until XSCF firmware reaches the ready state. This can be confirmed when the READY LEDs of XSCF_B#0 and XSCF_B#1 remain lit. 15.
XSCF> showlogs event Confirm no abnormality is found during the update. 20. Confirm that both the current and reserve banks of XSCFU#0 display the updated XCP versions. XSCF> version -c xcp XSCF#1 (Active ) XCP0 (Reserve): 1070 XCP1 (Current): 1070 XSCF#0 (Standby) XCP0 (Reserve): 1070 XCP1 (Current): 1070 If the Current and Reserve banks on XSCF#0 do not indicate XCP revision 1070, contact your authorized service representative. 21. Confirm that switching over between XSCF units works properly. a.
e. Confirm that XSCF#1 has entered the active state: XSCF> showlogs event .... Feb 26 16:10:28 PST 2008 XSCF#1 entered active state from standby state f. Confirm that no failures are found in system initialization: XSCF> showstatus No failures found in System Initialization. 22. Power on all domains. XSCF> poweron -a 23. Log in to XSCFU#0 and confirm all domains start up properly. XSCF> showlogs power 24. Check that there are no new errors.
Additional Software Procedures This section contains instructions for accomplishing some of the workarounds mentioned earlier in this document. Booting From a WAN Boot Server The WAN boot installation method enables you to boot and install software over a wide area network (WAN) by using HTTP. To support booting the Sun SPARC Enterprise M8000/M9000 server from a WAN boot server, you must have the appropriate wanboot executable installed to provide the needed hardware support.
krtld: load_exec: fail to expand cpu/$CPU krtld: error during initial load/link phase panic - boot: exitto64 returned from client program Identifying Degraded Memory in a System ▼ To Identify Degraded Memory in a System ● Log in to XSCF anf type the following command: XSCF> showstatus The following example identifies DIMM number 0A on Memory Board #5 has degraded memory.
▼ To Use the showdevices Command 1. Log in to XSCF and type the following command: XSCF> showdevices -d domain_id The following example shows that 00-0 has 64 Gbytes of memory while the other system boards have 16 Gbytes. XSCF> showdevices -d 0 ...
# prtdiag ...
XSCF> showdevices -d 0 ... Memory: ------DID 00 00 00 XSB 00-0 00-2 00-3 board mem MB 8192 8192 8192 perm mem MB 0 1674 0 base address 0x0000000000000000 0x000003c000000000 0x0000034000000000 domain target deleted remaining mem MB XSB mem MB mem MB 24576 24576 24576 ... The entry for column 4 perm mem MB indicates the presence of permanent memory if the value is not zero. The example shows permanent memory on 00-2, with 1674 Mbytes.
32 SPARC Enterprise M8000/M9000 Servers Product Notes for XCP 1070 • April 2008