Skip to main content

A.1.2 Configuration Procedure


A.1.2 Configuration Procedure
The following procedure updates a system that does not support physical partition dynamic reconfiguration to the one that does support it:
  1. Save logical domain configuration information to an XML file.
    If the PPAR DR function is enabled after the XCP firmware is updated, the logical domain configuration returns to the factory setting status (factory-default). So, the logical domain configuration information saved to XSCF cannot be used. For this reason, save the current logical domain configuration information to the XML file and then update the XCP firmware. After updating, restore the logical domain configuration information from the XML file to enable the easy reconfiguration of the logical domain.
    The following explains the procedure for saving the logical domain configuration information saved in the XSCF to an XML file.
    For details on the configuration information saved to the XML file, however, see the Oracle VM Server for SPARC Administration Guide.
  1. a. Log in to the control domain.

    b. Switch to the configuration information to be saved.

    Execute the ldm list-spconfig command on the control domain to list the logical domain configuration information. The line with [current] indicates the currently applied configuration information. In the following example, ldm-set3 is applied.
# ldm list-spconfig
factory-default
ldm-set1
ldm-set2
ldm-set3 [current]
  1. If only [next poweron] is displayed, there will be a difference between the logical domain configuration information stored in the XSCF and that stored in the control domain. So, execute the ldm add-spconfig command to save the current logical domain configuration information with a different name.
# ldm list-spconfig
factory-default
ldm-set1
ldm-set2
ldm-set3 [next poweron]
# ldm add-spconfig ldm-set4
# ldm list-spconfig
factory-default
ldm-set1
ldm-set2
ldm-set3
ldm-set4 [current]
  1. When the logical domain configuration information indicated in [current] matches the configuration information you want to save, proceed to step 1-c.

    If they are different, execute the ldm set-spconfig command to switch to the configuration information you want to save.

    In the following example, the configuration information is switched with the ldm set-spconfig command and then ldm-set1 is saved.
# ldm set-spconfig ldm-set1
# ldm list-spconfig
factory-default
ldm-set1 [next poweron]
ldm-set2
ldm-set3 [current]
  1. Execute the poweroff and poweron commands on XSCF to power off the physical partition (PPAR) and then power it back on. In this case, execute the poweroff command after changing the logical domain state to the state in which Oracle Solaris is running or the inactive state.

    The following example powers off PPAR-ID 0 and then powers it back on.
XSCF> poweroff -p 0
XSCF> poweron -p 0
  1. Execute the ldm list-spconfig command to check that the specified logical domain configuration information is set. The following example indicates that ldm-set1 is set as the current configuration information.
# ldm list-spconfig
factory-default
ldm-set1 [current]
ldm-set2
ldm-set3
  1. c. Execute the ldm list-constraints command to save the current logical domain configuration information to the XML file.

    Execute the ldm list-constraints command to save the current logical configuration information to the XML file.

    The following example saves the current logical domain configuration information to ldm-set1.xml. To prevent the saved XML file from being lost, back it up to another media or the like.
# ldm list-constraints -x > /ldm-set1.xml
  1. d. Check whether the configuration information is saved to the XML file.
    Execute the more command of Oracle Solaris or the like to check that the information has been saved to the XML file.
# more /ldm-set1.xml
<?xml version="1.0"?>
<LDM_interfaceversion="1.3" xmlns:xsi=http://www.w3.org/2001/XMLSchema-instancce
  1. e. To save any other logical domain configuration information, repeat steps 1.b to 1.e.

    If there is any other configuration information that you want to save, repeat steps 1.b to 1.e.
  1. Save the XSCF setting information.
    XSCF setting information can be saved to a USB device or to an external server through a network. The following describes each method.
  1. - Saving the XSCF setting information to a USB device
  1. a. Log in to the master XSCF.

    Execute the showbbstatus command to check that the XSCF to which you have logged in is the master XSCF. If the standby XSCF is being used, log in to the master XSCF again.
XSCF> showbbstatus
BB#00 (Master)
  1. b. Save the setting information to the USB device of the master XSCF.

    i. Connect the USB device to the USB port on the XSCF unit panel (rear panel) of the master XSCF.

    ii. Specify the output file name on the local USB device on the XSCF and then execute the dumpconfig command.

    The setting information is saved with the specified file name in base64 encoding text format.
XSCF> dumpconfig file:///media/usb_msd/backup-file.txt
operation completed
  1. When the "operation completed" message appears, data transfer has ended normally.
  1. iii. Once data transfer has ended, remove the USB device from the USB port.
  1. c. Using an editor or the like on a PC, check the following information saved in the setting file.
    - User-Comments: Comments made when the -c option was specified in the dumpconfig command

    - Created: Date and time when the information was saved

    - Platform: Model name

    - Serial-No: System serial number
XSCF Configuration File
User-Comments:
Encrypted: No
Created: Mon Jan 27 13:47:38 2014
Platform: M10-4S
Serial-No: 2111234001
Chassis-serial80:
Chassis-serial81:
Chassis-serial82:
Chassis-serial83:
Chassis-serial00:2111234001
Chassis-serial01:2111234003
Chassis-serial02:
Chassis-serial03:
Chassis-serial04:
Chassis-serial05:
Chassis-serial06:
Chassis-serial07:
Chassis-serial08:
Chassis-serial09:
Chassis-serial10:
Chassis-serial11:
Chassis-serial12:
Chassis-serial13:
Chassis-serial14:
Chassis-serial15:
Version: 0001
begin-base64_common
U1VOVyxTUEFSQy1FbnRlcnByaXNlAAAAAAAAAFLmZ6gAAPrfADhbdAAAAAIyMTExMjM0MDAzAAAA
...
  1. - Saving the setting information to an external server through a network
  1. a. Log in to the master XSCF.

    Execute the showbbstatus command to check that the XSCF to which you have logged in is the master XSCF. If the standby XSCF is being used, log in to the master XSCF again.
XSCF> showbbstatus
BB#00 (Master)
  1. b. Specify the target directory through a network to save the XSCF setting information.

    Specify the target directory and output file name and then execute the dumpconfig command. The setting information is saved with the specified file name in base64 encoding text format.
XSCF> dumpconfig -u "user-name" "ftp://server/backup/backup-sca-ff2-16.txt"
operation completed
  1. When the "operation completed" message appears, data transfer has ended normally.
  1. c. Using an editor or the like on a PC, check the saved setting file information.

    Check the following information.
    - User-Comments: Comments made when the -c option was specified in the dumpconfig command

    - Created: Date and time when the information was saved

    - Platform: Model name

    - Serial-No: System serial number
XSCF Configuration File
User-Comments:
Encrypted: No
Created: Mon Jan 27 13:47:38 2014
Platform: M10-4S
Serial-No: 2111234001
Chassis-serial80:
Chassis-serial81:
Chassis-serial82:
Chassis-serial83:
Chassis-serial00:2111234001
Chassis-serial01:2111234003
Chassis-serial02:
Chassis-serial03:
Chassis-serial04:
Chassis-serial05:
Chassis-serial06:
Chassis-serial07:
Chassis-serial08:
Chassis-serial09:
Chassis-serial10:
Chassis-serial11:
Chassis-serial12:
Chassis-serial13:
Chassis-serial14:
Chassis-serial15:
Version: 0001
begin-base64_common
U1VOVyxTUEFSQy1FbnRlcnByaXNlAAAAAAAAAFLmZ6gAAPrfADhbdAAAAAIyMTExMjM0MDAzAAAA
...
  1. Update Oracle Solaris and Oracle VM Server for SPARC.
    Obtain Oracle Solaris and Oracle VM Server for SPARC that support physical partition dynamic reconfiguration and apply them to the system.
  1. a. Update Oracle VM Server for SPARC of the control domain.

    See the latest Product Notes for your server to check the latest versions of Oracle Solaris and Oracle VM Server for SPARC. Then, update Oracle VM Server for SPARC using the following procedure.

    i. Obtain Oracle VM Server for SPARC for use with the control domain.

    - If the control domain is Oracle Solaris 10

    Obtain Oracle VM Server for SPARC that supports physical partition dynamic reconfiguration. For details on how to obtain this, see the latest Product Notes for your server.

    - If the control domain is Oracle Solaris 11

    Obtain SRU11.1.14.0 or later.

    ii. When Oracle VM Server for SPARC or Oracle Solaris for the control domain has been updated, restart the control domain.

    Restarting the control domain may cause an I/O domain to panic or a guest domain I/O to stop. If such a logical domain exists, stop it in advance with the shutdown command or stop it by executing the ldm stop-domain command from the control domain.

    Using the following method, you can check whether the logical domain is to be stopped.

    - For an I/O domain to which the PCle endpoint is assigned from the control domain

    Execute the ldm list-io command to determine whether the root complex (BUS) of the PCle endpoint assigned to the logical domain is assigned to primary. The following example indicates that the bus "PCIE2" of PCle endpoints "/BB0/PCI3" and "/BB0/PCI4" assigned to iodom0 is assigned to primary (control domain). Note that, since the configuration explained here differs from that in "A.1.1 Configuration Example," changes have been made to the logical domains and some other information.
# ldm list-io
NAME TYPE BUS DOMAIN STATUS
---- ---- --- ------ ------

PCIE0 BUS PCIE0 primary IOV
PCIE1 BUS PCIE1 primary IOV
PCIE2 BUS PCIE2 primary IOV
PCIE3 BUS PCIE3 primary IOV
....
/BB0/CMUL/NET0 PCIE PCIE0 primary OCC
/BB0/CMUL/SASHBA PCIE PCIE0 primary OCC
/BB0/PCI0 PCIE PCIE1 primary OCC
/BB0/PCI3 PCIE PCIE2 iodom0 OCC
/BB0/PCI4 PCIE PCIE2 iodom0 OCC
/BB0/PCI7 PCIE PCIE3 primary OCC
/BB0/PCI8 PCIE PCIE3 primary EMP
....
  1. - For the guest domain to which the virtual service of the control domain is assigned

    Execute ldm list-bindings primary to check the correspondence between the virtual network switch (VSW) and the connection destination (PEER), as well as between the virtual disk service (VDS) and the logical domain (CLIENT) that uses a virtual disk. In the following example, you can check that "guestdom0" is set for VSW PEER and VDS CLIENT.
# ldm list-bindings primary
....
VSW
NAME MAC NET-DEV ID DEVICE LINKPROP DEFAULT-VLAN-ID PVID
VID MTU MODE INTER-VNET-LINK
vsw0 00:14:4f:f9:88:ca net0 0 switch@0 1
1 1500 on
PEER MAC PVID VID MTU MAXBW LINKPROP
INTERVNETLINK
vnet0@guestdom0 00:14:4f:fa:64:dd 1 1500
VDS
NAME VOLUME OPTIONS MPGROUP DEVICE
....
CLIENT VOLUME
vdisk0@guestdom0 vol0
  1. Execute the ldm stop-domain command to stop the logical domain checked above. In the following example, the ldm stop-domain command is executed to stop iodom0 and guestdom0 and then the ldm list-domain command is executed to check that the bound state is set.
# ldm stop-domain guestdom0
LDom guestdom0 stopped
# ldm stop-domain iodom0
LDom iodom0 stopped
# ldm list-domain
NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME
primary active -n-cv- UART 64 58G 0.0% 6h 3m
guestdom0 bound ------ 5100 64 64G
iodom0 bound ----v- 5000 32 32G
  1. iii. Update Oracle VM Server for SPARC of the control domain.

    - For Oracle Solaris 10

    Uninstall the old Oracle VM Server for SPARC and then install the new Oracle VM Server for SPARC. For details, see the README file provided with the obtained Oracle VM Server for SPARC.

    - For Oracle Solaris 11

    Apply the SRU. For details, see the installation manual of the obtained SRU.

    iv. Execute the shutdown command to restart the control domain.
# shutdown -i6 -g0 -y
....
  1. v. Execute the ldm start-domain command to start the logical domain stopped in ii, above.

    In the following example, the ldm start-domain command is executed to start "iodom0" and "guestdom0" and then the ldm list-domain command is executed to check that the active state is set.
# ldm start-domain guestdom0
# ldm start-domain iodom0
# ldm list-domain
NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME
primary active -n-cv- UART 64 58G 0.0% 6h 3m
guestdom0 active -n---- 5100 64 64G
iodom0 active -n--v- 5000 32 32G
  1. b. Update Oracle Solaris for logical domains other than the control domain.

    See the latest Product Notes for your server, and update Oracle Solaris for logical domains other than the control domain. For details on the update procedure, see the information relating to each update.
  1. Update the XCP firmware.
    Install the PSB (BB) mounted on the SPARC64 X+ processor in the physical partition configured with the SPARC64 X processor. To do this, update the physical partition configured with the SPARC64 X processor to the XCP firmware of XCP 2220 or later.
    a. Obtain the latest XCP firmware.

    i. Download the program file for the XCP firmware.

    Download, from the web site, the program file (XCPxxxx.tar.gz or XCPxxxx.exe) for the XCP firmware to any folder on a PC connected to this system.

    Use one of the following methods to obtain the firmware for the server you are using.

    - Japanese site

    A customer who has entered into a contract with the SupportDesk can obtain firmware from SupportDesk-Web.

    - Global site

    For details on the method used to obtain the latest file of the firmware, contact our sales personnel.

    The following files are provided.

    - Firmware program file
    (XSCF Control Package (XCP) file)

    - XSCF extended MIB (XSCF-SP-MIB) definition file

    ii. Check the XCP version of the downloaded program file.

    Check the version of the program file of the downloaded XCP firmware. For the XCP version, refer to the 4-digit number in the file name of the firmware program (tar.gz format) to check that it is the XCP firmware version update to be applied. For example, if the name of the program file is "XCP2220.tar.gz," the XCP version is 2220.

    iii. Decompress the downloaded program file.

    Decompress the program file of the downloaded XCP firmware. The XCP image file to import into the system is expanded. For example, if "XCP2220.tar.gz" is decompressed, "BBXCP2220.tar.gz" is expanded.

    b. Check the current XCP firmware version.

    i. Log in to the master XSCF.
    Execute the showbbstatus command to check that the XSCF to which you have logged in is the master XSCF. If the standby XSCF is being used, log in to the master XSCF again.
XSCF> showbbstatus
BB#00 (Master)
  1. ii. Execute the version command to check the XCP version of the current system.

    Before updating the firmware, check the XCP version of the current system. In the following example, the version command is executed with the -c xcp option added, to check if the XCP version is XCP 2041 that does not support physical partition dynamic reconfiguration.
XSCF> version -c xcp
BB#00-XSCF#0 (Master)
XCP0 (Current): 2041
XCP1 (Reserve): 2041
BB#01-XSCF#0(Standby)
XCP0 (Current): 2041
XCP1 (Reserve): 2041
  1. c. Update the XCP firmware.

    i. Execute the poweroff command to power off all the physical partitions.
XSCF> poweroff -a
  1. Execute the showpparstatus command to check that all the physical partitions are powered off.
XSCF> showpparstatus -a
PPAR-ID PPAR Status
0 Powered Off
  1. ii. Execute the showhardconf command to check that [Status] of the master and standby XSCFs is "Normal."
XSCF> showhardconf
SPARC M10-4S;
+ Serial: 2081230011; Operator_Panel_Switch:Service;
+ System_Power:On; System_Phase:Cabinet Power On;
Partition#0 PPAR_Status:Running;
BB#00 Status:Normal; Role:Master; Ver:2003h; Serial:2081231002;
+ FRU-Part-Number: CA07361-D202 A1 ;
+ Power_Supply_System: ;
+ Memory_Size:256 GB;
  :
BB#01 Status:Normal; Role:Standby; Ver:0101h;Serial:7867000297;
+ FRU-Part-Number: CA20393-B50X A2 ;
+ Power_Supply_System: ;
+ Memory_Size:256 GB;
CMUL Status:Normal; Ver:0101h; Serial:PP123406CB ;
+ FRU-Part-Number:CA07361-D941 C4 /7060911 ;
+ Memory_Size:128 GB;
   :
  1. iii. Execute the getflashimage command to import the XCP image file.

    In the following example, the USB device is connected to the USB port (MAINTENANCE ONLY printed) on the XSCF unit panel (rear panel) of the master XSCF and the XCP image file is imported.
XSCF> getflashimage file:///mnt/share/scf-firm/xscf/user/scfadmin/BBXCP2220.tar.gz
0MB received
1MB received
2MB received
...
86MB received
87MB received
88MB received
Download successful: 90539 Kbytes in 58 secs (1562.668 Kbytes/sec) Checking file...
MD5: 2b89c06548205ce35a8ecb6c2321d999
  1. When the normal end messages "Download successful: ..." and "MD5: ..." appear, they indicate that the importing of the XCP image file has ended.
Note - The message "Warning: About to delete existing old versions." may appear when the XCP image file is imported. This message asks you to confirm whether to delete an old XCP image file that has already been imported. If "Continue?" appears, enter "y" to continue import processing.
Note - If the message "Error:File is invalid or corrupt" appears after the XCP image file is imported, this indicates that the imported XCP image file is inappropriate. The XCP image file may have been destroyed. So, obtain the correct XCP image file and then import it.
  1. iv. Execute the getflashimage -l command to check the version of the imported XCP image file.
XSCF> getflashimage -l
Existing versions:
Version Size Date
BBXCP2220.tar.gz 92712351 Thu May 23 15:01:42 JST 2014
  1. v. Execute the flashupdate -c check command to check whether the imported XCP image file can be used for update.

    Execute the showresult command immediately after the execution of the flashupdate command. If the end value is 0, update is possible.
XSCF> flashupdate -c check -m xcp -s 2220
XCP update is started. [3600sec]
0XSCF>
XSCF> showresult
0
XSCF>
  1. vi. Execute the flashupdate command to update the firmware.
XSCF> flashupdate -c update -m xcp -s 2220
The XSCF will be reset. Continue? [y|n]: y
XCP update is started. [3600sec]
0.....30.....60.....90.....120.....150.....180.....210.....240.....-

270.....300.....330.....360.....390.....420.....450.....480.....510.....|

540.....570.....600.....630.....660.....690.....720.....750.....780.....-

810.....840.....870.....900.....930

:
  1. Here, the XSCF is rebooted and the XSCF session is disconnected. At this point, the XCP firmware update has not yet been completed.
Note - Regarding the work time for the firmware, the update takes about 60 minutes and the automatic switching of the XSCF after the completion of the update takes about 10 minutes.
Note - To perform the update safely, do not operate the power supply of the physical partition until the message "XCP update has been completed" appears, to indicate that the XCP firmware has been updated.
  1. vii. Connect to the master XSCF again.

    Immediately after an XSCF reboot, the master and standby XSCFs will be the exact reverse of the original states. For example, if the firmware update is executed with the master XSCF of BB-ID 0 and a connection is then made to the XSCF again, BB-ID 1 enters the master state and BB-ID 0 enters the standby state.
Note - If an inherited IP address is set and is used for connection, a connection is automatically made to the master XSCF.
  1. viii. Execute the showbbstatus command to check that you have logged in to the master XSCF. If the standby XSCF is being used, connect to the master XSCF again.
XSCF> showbbstatus
BB#01 (Master)
  1. ix. Execute the showlogs monitor command to check that the XCP firmware update has been completed.

    If the message "XCP update has been completed" appears, XCP firmware update has been completed.
XSCF> showlogs monitor
May 29 09:38:05 M10-4S-0 Event: SCF:XCP update is started (XCP version=2220:
last version=2041)
May 29 09:40:31 M10-4S-0 Event: SCF:XSCF update is started (BBID=0, bank=0)
May 29 09:40:46 M10-4S-0 Event: SCF:XSCF update is started (BBID=1, bank=0)
May 29 09:41:03 M10-4S-0 Event: SCF:XSCF writing is performed (BBID=0, XSCF
version=02220000)
:
May 29 09:50:41 M10-4S-0 Event: SCF:XSCF bank apply has been completed
(BBID=1, bank=0, XCP version=2220: last version=2041)
May 29 10:01:50 M10-4S-0 Event: SCF:XSCF bank apply has been completed
(BBID=0, bank=0, XCP version=2220: last version=2041)
May 29 10:01:51 M10-4S-0 Event: SCF:Change Master Start (BB#01)May 29 10:03:26 M10-4S-1 Event: SCF:Change Master Complete (BB#01)
May 29 10:05:00 M10-4S-1 Event: SCF:Standby XSCF Ready (BBID#00)
:
May 29 10:32:38 M10-4S-1 Event: SCF:XCP update has been completed (XCP
version=2220: last version=2041)
May 29 10:32:39 M10-4S-1 Event: SCF:This XSCF will be switched back to standby
mode after completion of firmware update
May 29 10:32:39 M10-4S-1 Event: SCF:Change Master Start (BB#00)
May 29 10:33:29 M10-4S-1 Event: SCF:Change Master Complete (BB#00)
May 29 10:42:29 M10-4S-1 Event: SCF:Standby XSCF Ready (BBID#01)
Subsequently, the master and standby states of the XSCF are automatically switched.

Switching has been completed when the messages "This XSCF will be switched back to standby mode after completion of firmware update", "Change Master Complete", and "Stand by XSCF Ready" appear.
Note - If the message "XCP update has been completed" does not appear, update has not yet been completed. Execute the showlogs monitor command again to check that update has been completed.
Note - If firmware update is executed, the master and standby states are switched twice. As a result, the master XSCF returns to the XSCF for which the flashupdate command was executed. When the master and standby states are switched, the XSCF session may be disconnected. If the XSCF session has been disconnected, connect to it again.
To check that switching has been completed, execute the showhardconf command to check that [Status] of the master and standby XSCFs is "Normal."
XSCF> showhardconf
SPARC M10-4S;
+ Serial: 2081230011; Operator_Panel_Switch:Service;
+ System_Power:On; System_Phase:Cabinet Power On;
Partition#0 PPAR_Status:Running;
BB#00 Status:Normal; Role:Master; Ver:2003h; Serial:2081231002;
+ FRU-Part-Number: CA07361-D202 A1 ;
+ Power_Supply_System: ;
+ Memory_Size:256 GB;
  :
BB#01 Status:Normal; Role:Standby; Ver:0101h;Serial:7867000297;
+ FRU-Part-Number: CA20393-B50X A2 ;
+ Power_Supply_System: ;
+ Memory_Size:256 GB;
CMUL Status:Normal; Ver:0101h; Serial:PP123406CB ;
+ FRU-Part-Number:CA07361-D941 C4 /7060911 ;
+ Memory_Size:128 GB;
   :
Note - Suppose that the showhardconf command is executed and the message "Cannot communicate with the other XSCF. Check the other XSCF's state." appears. This indicates that switching between the master and standby states of the XSCF has not yet been completed. Execute the showhardconf command again to check that switching has been completed.
  1. x. To return the master and standby XSCFs to the states they had before the update, execute the switchscf command.
XSCF> switchscf -t Standby
The XSCF unit switch between the Active and Standby states.
Continue? [y|n] :y
  1. Since the XSCF session is disconnected by an XSCF reboot, connect to the master XSCF again.

    To check that switching has been completed, execute the showhardconf command to check that [Status] of the master and standby XSCFs is "Normal."
Note - Suppose that the showhardconf command is executed and the message "Cannot communicate with the other XSCF. Check the other XSCF's state." appears. This indicates that switching between the master and standby states of the XSCF has not yet been completed. Execute the showhardconf command again to check that switching has been completed.
  1. xi. Execute the version command to check that the firmware is the new version.
XSCF> version -c xcp
BB#00-XSCF#0 (Master)
XCP0 (Current): 2220
XCP1 (Reserve): 2220
BB#01-XSCF#0(Standby)
XCP0 (Current): 2220
XCP1 (Reserve): 2220
Note - If the physical partition is powered on and the firmware is updated, the current bank of the CMU firmware will be new. If the physical partition is powered off and the firmware is updated, both the reserve bank and current bank of the CMU firmware will be new. For details on the versions of the CMU firmware associated with the XCP versions, see "Existing XCP Firmware Versions and Support Information" in the latest product notes.
  1. Specify the XSCF PPAR DR mode and logical domain to be started, and then start the physical partition.
Note - To use physical partition dynamic reconfiguration, it is necessary to enable the PPAR DR mode using the setpparmode command. However, if the PPAR DR mode is enabled, the logical domain configuration information is changed to the factory-default. So, the existing logical domain configuration information cannot be used. In this case, you must create new logical domain configuration information.
To use the existing logical domain configuration information, disable the PPAR DR mode. Suppose that the PPAR DR mode is enabled and that the power to the physical partition is turned on while the existing logical domain configuration information is used. Problems such as a hypervisor abort and a failure to start Oracle VM Server for SPARC normally will occur.
  1. a. Enable the physical partition dynamic reconfiguration function (PPAR DR function).

    i. Execute the showpparmode command to reference the PPAR DR mode setting.

    The following example specifies physical partition number 0 (PPAR ID 0).
XSCF> showpparmode -p 0
Host-ID :9007002b
Diagnostic Level :min
Message Level :normal
Alive Check :on
Watchdog Reaction :reset
Break Signal :on
Autoboot(Guest Domain) :on
Elastic Mode :off
IOreconfigure :false
PPAR DR(Current) :-
PPAR DR(Next) :off
  1. ii. If the PPAR DR mode is disabled, execute the setpparmode command to enable that mode.
XSCF> setpparmode -p 0 -m ppar_dr=on
  1. iii. Execute the showpparmode command to check that the PPAR DR mode has been enabled.
XSCF> showpparmode -p 0
Host-ID :9007002b
Diagnostic Level :min
Message Level :normal
Alive Check :on
Watchdog Reaction :reset
Break Signal :on
Autoboot(Guest Domain) :on
Elastic Mode :off
IOreconfigure :false
PPAR DR(Current) :-
PPAR DR(Next) :on
  1. b. Power on the physical partition.

    i. Execute the showdomainconfig command to check that the configuration information for the logical domains to be started is set to the factory-default.

    The following example specifies physical partition number 0 (PPAR ID 0).
XSCF> showdomainconfig -p 0
PPAR-ID :0
Booting config
(Current) : factory-default
(Next) : factory-default
----------------------------------------------------------------

Index :1
config_name :factory-default
domains :1
date_created:-
----------------------------------------------------------------

Index :2
config_name :ldm-set1
domains :8
date_created:'2012-08-0811:34:56'
----------------------------------------------------------------

Index :3
config_name :ldm-set2
domains :20
date_created:'2012-08-0912:43:56':
  1. ii. Execute the poweron command to power on the physical partitions.

    The following example specifies physical partition number 0 (PPAR ID 0).
XSCF> poweron -p 0
  1. Using the XML file, reconfigure the logical domain.

    After completing the startup of the physical partition, reconfigure the logical domain from the XML file saved in step 1.
    a. Check that the system has started with the factory-default settings for the control domain.

    Execute the ldm list-spconfig command to check that [current] is displayed on the factory-default line.
# ldm list-spconfig
factory-default [current]
ldm-set1
ldm-set2
ldm-set3
  1. b. Using the ldm init-system command, apply the logical domain configuration information from the XML file.

    Execute the ldm init-system command to apply the saved XML file setting. Then, execute the shutdown command to restart the control domain.
# ldm init-system -i /ldm-set1.xml
Initiating a delayed reconfiguration operation on the primary domain.
All configuration changes for other domains are disabled until the primary
domain reboots, at which time the new configuration for the primary domain
will also take effect.
# shutdown -y -g0 -i6
  1. c. To enable physical partition dynamic reconfiguration, adjust the memory size of each logical domain.

    Execute the ldm list-domain command or the like to check that the logical domains have been configured from the XML file.
# ldm list-domain
NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME
primary active -n-cv- UART 64 58G 0.0% 6m
guest0 inactive ------ 64 64G
guest1 inactive ------ 64 64G
root-dom0 inactive ------ 32 32G
root-dom1 inactive ------ 32 32G
  1. To enable physical partition dynamic reconfiguration when the version of Oracle VM Server for SPARC is earlier than 3.2, set the memory size to be assigned to the logical domain to a multiple of "number of cores assigned to the logical domain x 256 MB". To re-set the memory size of a logical domain other than the control domain, use the ldm set-memory command.

    To re-set the memory size of the control domain, do so as follows: Enter delayed reconfiguration mode with the ldm start-reconf command, set the same number of cores with the ldm set-core command, and then re-set the memory size with the ldm set-memory command. Then, restart Oracle Solaris.

    When the version of Oracle VM Server for SPARC is earlier than 3.2, first, obtain a multiple of 8192 MB to make the value closer to the original setting value (58 GB = 59392 MB). The result is 59392/8192 = 7.25. So, round it down to 7. As a result, the memory size to be reassigned to the control domain is 81924 MB × 7 = 56 GB.

    Next, the following example shows the execution of the command for re-setting the control domain memory to 56 GB.
# ldm start-reconf primary
Initiating a delayed reconfiguration operation on the primary domain.
All configuration changes for other domains are disabled until the primary
domain reboots, at which time the new configuration for the primary domain
will also take effect.
# ldm set-core 32 primary
------------------------------------------------------------------------------

Notice: The primary domain is in the process of a delayed reconfiguration.
Any changes made to the primary domain will only take effect after it reboots.
------------------------------------------------------------------------------

# ldm set-memory 56G primary
------------------------------------------------------------------------------

Notice: The primary domain is in the process of a delayed reconfiguration.
Any changes made to the primary domain will only take effect after it reboots.
------------------------------------------------------------------------------

# shutdown -i6 -g0 -y
....
  1. Execute the ldm list-domain command to check that the memory size (MEMORY) has been set properly.
# ldm list-domain
NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME
primary active -n-cv- UART 64 56G 0.0% 6m
guest0 inactive ------ 64 64G
guest1 inactive ------ 64 64G
root-dom0 inactive ------ 32 32G
root-dom1 inactive ------ 32 32G
  1. Execute the ldm bind-domain command to bind each logical domain.
# ldm bind-domain root-dom0
# ldm bind-domain root-dom1
# ldm bind-domain guest0
# ldm bind-domain guest1
  1. However, suppose that the logical domain configuration information is restored from an XML file that was saved when the PPAR DR function was disabled. In this case, the memory size of the hypervisor area increases when the PPAR DR function is enabled, such that the amount of allocated memory will be insufficient. If the ldm bind-domain command is executed while insufficient memory is allocated, the processing will end with an error together with the following error message, and the inactive state remains as is.
# ldm bind-domain guest1
Not enough free memory present to meet this request.
Could not bind requested memory for LDom guest1.
  1. In this case, see step 6.c to re-set the memory size allocated to the logical domains. To adjust the memory allocation of a logical domain that has already been placed in the bound state, execute the ldm unbind-domain command to release the bound state and then make the setting. When re-setting has been completed, re-execute the ldm bind-domain command for re-binding.
  1. d. Check the memory resource allocation status.

    Execute the ldm list-devices –a command to check the status of the continuous area (memory block) of the memory allocated to each logical domain. Then, check that there are not a large number of small memory blocks. As a guide, check that there is not a large number of memory blocks with a SIZE of about 256 MB to 512 GB.
# ldm list-devices -a memory
MEMORY
PA SIZE BOUND
0x700000000000 32G root-dom0
0x720000000000 32G root-dom1
0x740000000000 32G guest0
0x760000800000 1272M _sys_
0x760050000000 2G primary
0x7600d0000000 29440M guest1
0x780000000000 32G guest0
0x7a0000000000 32G guest1
0x7c0000000000 32G primary
0x7e0000800000 1272M _sys_
0x7e0050000000 512M _sys_
0x7e0070000000 256M _sys_
0x7e0080000000 3328M guest1
0x7e0150000000 4864M
0x7e0280000000 22G primary
  1. If the memory block is broken up, execute the ldm unbind-domain command to restore each logical domain to the unbind status. Then, perform a re-set as described in step 6.c.
  1. e. Start all the logical domains.

    Execute the ldm start-domain command to start the logical domains other than the control domain. Suppose that there are dependency relationships such as a physical I/O allocation relationship (root domain, I/O domain) and relationships between virtual I/O services and clients. In this case, first start the supply source domain of the physical and virtual I/O services.

    The following example shows that domains are started in the order assumed for the configuration example. The ldm start-domain command is used to start root-dom0, root-dom1, guest0, and guest1, in that order, and the ldm list-domain command is used to check that these domains are started.  If [STATE] is "active" and the second character from the left of the FLAGS string is "n", this indicates that Oracle Solaris has started.
# ldm start-domain root-dom0
LDom root-dom0 started
# ldm start-domain root-dom1
LDom root-dom1 started
# ldm start-domain guest0
LDom guest0 started
# ldm start-domain guest1
LDom guest1 started
# ldm list-domain
NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME
primary active -n-cv- UART 64 56G 0.2% 15m
guest0 active -n---- 5002 64 64G 0.4% 18s
guest1 active -n---- 5003 64 64G 7.9% 9s
root-dom0 active -n--v- 5000 32 32G 0.5% 24s
root-dom1 active -n--v- 5001 32 32G 1.6% 16s
  1. f. Save the logical domain configuration information to the XSCF.

    After restoring the logical domain configuration information, execute the ldm add-spconfig command to save the restored logical domain configuration information to the XSCF. The following example saves the ldm-set4 configuration information to the XSCF.
# ldm add-spconfig ldm-set4
# ldm list-spconfig
factory-default
ldm-set1
ldm-set2
ldm-set3
ldm-set4 [current]
  1. g. If there are multiple items of logical domain configuration information to be restored, restore the factory-default configuration and then repeat steps 6.b to 6.g.

    The following example shows the return to the factory-default configuration.
    Execute the showdomainconfig command from the XSCF to check the configuration information for the logical domain to be started.
XSCF> showdomainconfig -p 0
PPAR-ID :0
Booting config
(Current) :ldm-set1
(Next) :ldm-set1
----------------------------------------------------------------

Index :1
config_name :factory-default
domains :1
date_created:-
----------------------------------------------------------------

Index :2
config_name :ldm-set1
domains :8
date_created:'2012-08-0811:34:56'
----------------------------------------------------------------

....
  1. Execute the setdomainconfig command to specify factory-default Index1 in physical partition number 0 (PPAR ID 0).
XSCF> setdomainconfig -p 0 -i 1
  1. Execute the poweroff and then the poweron commands to restart the physical partition. Once startup is complete, repeat steps 6.b to 6.g.
XSCF> poweroff -p 0
...
XSCF> poweron -p 0
...