Skip to main content

Contents

  1. Preface
  2. Chapter 1 Understanding an Overview of the SPARC M12/M10
    1.  1.1 Basics of the SPARC M12/M10
    2. Click 1.2 Basics of the XSCF Firmware
    3. Click 1.3 Network Configuration
    4.  1.4 Basics of Hypervisor
    5.  1.5 Basics of Oracle VM Server for SPARC
    6.  1.6 Basics of OpenBoot PROM
  3. Chapter 2 Logging In/Out of the XSCF
    1. Click 2.1 Connecting the System Management Terminal
    2. Click 2.2 Logging In to the XSCF Shell
    3.  2.3 Logging Out from the XSCF Shell
    4. Click 2.4 Logging In to XSCF Web
    5.  2.5 Logging Out From XSCF Web
    6.  2.6 Number of Connectable Users
  4. Chapter 3 Configuring the System
    1. Click 3.1 Preliminary Work for XSCF Setup
    2. Click 3.2 Understanding the Contents of the XSCF Firmware Settings
    3.  3.3 Setting Up From the XSCF Shell
    4.  3.4 Setting Up From XSCF Web
    5. Click 3.5 Creating/Managing XSCF Users
    6. Click 3.6 Setting the XSCF Time/Date
    7. Click 3.7 Configuring the SSH/Telnet Service for Login to the XSCF
    8. Click 3.8 Configuring the HTTPS Service for Login to the XSCF
    9. Click 3.9 Configuring the XSCF Network
    10. Click 3.10 Configuring Auditing to Strengthen XSCF Security
  5. Chapter 4 Configuring the System to Suit the Usage Type
    1. Click 4.1 Setting/Checking the System Altitude
    2. Click 4.2 Controlling System Start
    3. Click 4.3 Enabling/Disabling Dual Power Feed
    4. Click 4.4 Reducing Power Consumption
    5. Click 4.5 Connecting a DVD Drive
    6. Click 4.6 Using Remote Storage
  6. Chapter 5 CPU Activation
    1.  5.1 Basic Concepts of CPU Activation
    2.  5.2 CPU Activation Key
    3. Click 5.3 Adding CPU Core Resources
    4. Click 5.4 Deleting CPU Core Resources
    5. Click 5.5 Moving CPU Core Resources
    6. Click 5.6 Displaying CPU Activation Information
    7. Click 5.7 Saving/Restoring CPU Activation Keys
    8. Click 5.8 Troubleshooting CPU Activation Errors
    9.  5.9 Important Notes about CPU Activation
  7. Chapter 6 Starting/Stopping the System
    1. Click 6.1 Starting the System
    2. Click 6.2 Stopping the System
    3.  6.3 Rebooting the System
    4.  6.4 Suppressing Starting Oracle Solaris at Power-on
  8. Chapter 7 Controlling Physical Partitions
    1.  7.1 Configuring a Physical Partition
    2. Click 7.2 Setting the Physical Partition Operation Mode
    3.  7.3 Powering On a Physical Partition
    4.  7.4 Powering Off a Physical Partition
    5.  7.5 Changing the Configuration of a Physical Partition
  9. Chapter 8 Controlling Logical Domains
    1.  8.1 Configuring a Logical Domain
    2. Click 8.2 Configuring the Oracle Solaris Kernel Zone
    3. Click 8.3 Switching to the Control Domain Console From the XSCF Shell
    4. Click 8.4 Returning to the XSCF Shell From the Control Domain Console
    5.  8.5 Starting a Logical Domain
    6.  8.6 Shutting Down a Logical Domain
    7. Click 8.7 Ordered Shutdown of Logical Domains
    8.  8.8 Checking CPU Activation Information
    9. Click 8.9 Setting the OpenBoot PROM Environment Variables of the Control Domain
    10. Click 8.10 Domain Console Logging Function
    11.  8.11 Changing the Configuration of a Logical Domain
    12.  8.12 Setting the Logical Domain Time
    13. Click 8.13 Collecting a Hypervisor Dump File
    14. Click 8.14 Managing Logical Domain Resources Associated with CPU Sockets
    15. Click 8.15 Setting the Physical Partition Dynamic Reconfiguration Policy
    16. Click 8.16 Setting the Maximum Page Size of a Logical Domain
  10. Chapter 9 Managing the Systems Daily
  11. Chapter 10 Preparing/Taking Action for Failures
    1.  10.1 Learning about Troubleshooting and Related Functions
    2. Click 10.2 Receiving Notification by E-mail When a Failure Occurs
    3. Click 10.3 Monitoring/Managing the System Status With the SNMP Agent
    4. Click 10.4 Monitoring the System
    5.  10.5 Understanding the Failure Degradation Mechanism
    6. Click 10.6 Checking Failed Hardware Resources
    7. Click 10.7 Setting Automatic Replacement of Failed CPU Cores
    8.  10.8 Setting Recovery Mode
    9.  10.9 Setting Up a Redundant Component Configuration
    10. Click 10.10 Saving/Restoring XSCF Settings Information
    11. Click 10.11 Saving/Restoring Logical Domain Configuration Information in the XSCF
    12. Click 10.12 Saving/Restoring Logical Domain Configuration Information in an XML File
    13. Click 10.13 Saving/Restoring the OpenBoot PROM Environment Variables
    14.  10.14 Saving/Restoring the Contents of a Hard Disk
    15.  10.15 Resetting a Logical Domain
    16. Click 10.16 Causing a Panic in a Logical Domain
    17.  10.17 Resetting a Physical Partition
    18. Click 10.18 Returning the Server to the State at Factory Shipment
    19.  10.19 Collecting a Crash Dump File Using Deferred Dump
  12. Chapter 11 Checking the System Status
    1. Click 11.1 Checking the System Configuration/Status
    2. Click 11.2 Checking a Physical Partition
  13. Chapter 12 Checking Logs and Messages
    1. Click 12.1 Checking a Log Saved by the XSCF
    2. Click 12.2 Checking Warning and Notification Messages
  14. Chapter 13 Switching to Locked Mode/Service Mode
    1.  13.1 Understanding the Differences Between Locked Mode and Service Mode
    2.  13.2 Switching the Operating Mode
  15. Chapter 14 Configuring a Highly Reliable System
    1. Click 14.1 Configuring Memory Mirroring
    2. Click 14.2 Configuring Hardware RAID
    3.  14.3 Using the LDAP Service
    4.  14.4 Using SAN Boot
    5.  14.5 Using iSCSI
    6. Click 14.6 Remote Power Management for the SPARC M12/M10 and I/O Devices
    7.  14.7 Using an Uninterruptible Power Supply
    8. Click 14.8 Using Verified Boot
  16. Chapter 15 Expanding the System Configuration
    1.  15.1 Changing the Virtual CPU Configuration
    2.  15.2 Changing the Memory Configuration
    3. Click 15.3 Dynamic Reconfiguration Function for PCIe Endpoint Devices
    4. Click 15.4 Using the PCI Expansion Unit
    5.  15.5 Expanding the SPARC M12-2S/M10-4S
  17. Chapter 16 Updating the XCP Firmware
    1. Click 16.1 Basics of Firmware Update
    2. Click 16.2 Before Updating Firmware
    3.  16.3 Update Flow
    4.  16.4 Preparing an XCP Image File
    5. Click 16.5 Updating Firmware
    6.  16.6 Updating Firmware From XSCF Web
    7. Click 16.7 Firmware Version Matching with Parts Addition/Replacement
    8.  16.8 Trouble During Firmware Update
    9.  16.9 FAQ Relating to Firmware Update
  18. Chapter 17 Updating Oracle Solaris and Oracle VM Server for SPARC
  19. Chapter 18 Troubleshooting
    1.  18.1 Troubleshooting for the XSCF
    2.  18.2 Precautions Concerning Using the RESET Switch
    3.  18.3 Frequently Asked Questions / FAQ
    4.  18.4 System Troubleshooting With the XSCF
  20. Appendix A Lists of SPARC M12/M10 System Device Paths
    1.  A.1 SPARC M12-1 Device Paths
    2. Click A.2 SPARC M12-2 Device Paths
    3. Click A.3 SPARC M12-2S Device Paths
    4.  A.4 SPARC M10-1 Device Paths
    5. Click A.5 SPARC M10-4 Device Paths
    6. Click A.6 SPARC M10-4S Device Paths
  21. Appendix B Identifying an SAS2 Device Based on a WWN
    1.  B.1 World Wide Name (WWN) Syntax
    2.  B.2 Overview of probe-scsi-all Command Output
    3. Click B.3 Identifying a Disk Slot by Using the probe-scsi-all Command
    4. Click B.4 Identifying Disk Slot
  22. Appendix C List of the XSCF Web Pages
    1.  C.1 Overview of the Pages
    2.  C.2 Understanding the Menu Configuration
    3. Click C.3 Available Pages
  23. Appendix D XSCF MIB Information
    1.  D.1 MIB Object Identification
    2.  D.2 Standard MIB
    3. Click D.3 Extended MIB
    4.  D.4 Traps
  24. Appendix E SPARC M12/M10 System-specific Functions of Oracle VM Server for SPARC
    1.  E.1 Ordered Shutdown of Logical Domains
    2.  E.2 CPU Activation Support
    3. Click E.3 Checking Failed Resources
    4.  E.4 Automatic Replacement of Failed CPUs
    5.  E.5 Hypervisor Dump
    6.  E.6 Domain Console Logging Function
    7.  E.7 CPU Socket Restrictions
  25. Appendix F SAS2IRCU Utility Command Examples
    1.  F.1 Displaying a List of SAS Controllers Recognized by sas2ircu
    2.  F.2 Displaying the Information of a Hardware RAID Volume
    3.  F.3 Adding a Hardware RAID Volume
    4.  F.4 Displaying the Configuration Status of a Hardware RAID Volume
    5.  F.5 Creating a Hot Spare of a Hardware RAID Volume
    6.  F.6 Deleting a Hot Spare of a Hardware RAID Volume
    7.  F.7 Deleting a Hardware RAID Volume
    8.  F.8 Identifying the Faulty Disk Drive of a Hardware RAID Volume
  26. Appendix G SPARC M12-1/M10-1 XSCF Startup Mode Function
    1. Click G.1 Function Overview
    2. Click G.2 Restrictions and Notes
    3.  G.3 Configuration Procedure
  27. Appendix H OpenBoot PROM Environment Variables and Commands
    1.  H.1 SCSI Device Display
    2.  H.2 Unsupported OpenBoot PROM Environment Variables
    3.  H.3 Unsupported OpenBoot PROM Commands
    4.  H.4 Behavior With the Security Mode Enabled
  28. Appendix I How to Specify the Boot Device
    1.  I.1 Device Path of the Internal Storage
    2.  I.2 Method of Specification With a PHY Number
    3.  I.3 Method of Specification With a Target ID
    4.  I.4 Method of Specification With an SAS Address
    5.  I.5 Method of Specification With a Volume Device Name
    6.  I.6 Notes About the Device Alias net of the SPARC M12 Without On-Board LAN
  29. Appendix J Lists of DVD Drive Aliases
    1. Click J.1 External DVD Drive Aliases
    2. Click J.2 Remote Storage DVD Drive Aliases
  30. Appendix K CPU Activation Interim Permit
    1.  K.1 What is the CPU Activation Interim Permit?
    2.  K.2 Terms of Use of the CPU Activation Interim Permit and Precautions
    3. Click K.3 Related Commands
    4. Click K.4 Flows and Procedures for Using a CPU Activation Interim Permit
    5. Click K.5 Event Notification of the CPU Activation Interim Permit
    6. Click K.6 Other Important Notes

K.4.3 Case Where the Function Has Expired or is Disabled


K.4.3 Case Where the Function Has Expired or is Disabled
This section describes system operation and action to be performed when a CPU Activation Interim Permit expires or is disabled.
Configuration Example
The description here uses the following configuration as an example.
  1. Single CPU socket SPARC M10-1 with 16 physical CPU cores is used.
  1. CPU Activations to activate 4 CPU cores have been purchased and registered in the system.
  1. The system consists of two logical domains and they are running.
  1. - Two CPU cores are assigned to the control domain (primary domain).
  1. - Two CPU cores are assigned to the guest domain (dm0).
  1. To respond to a load increase, the guest domain (dm0) requires six more CPU cores.
  1. A CPU Activation Interim Permit is enabled, and six additional CPU cores have been assigned to the guest domain (dm0) as a temporary action.
Figure K-3  Configuration Example for the Case Where the Function Has Expired or is Disabled (SPARC M10-1)
Figure K-3 Configuration Example for the Case Where the Function Has Expired or is Disabled (SPARC M10-1)
Note - CPU Activation keys registered in the system enable a definite number of CPU cores. They do not enable CPU cores by specifying the CPU socket and physical location of each CPU core. For the purpose of easy understanding, specific CPU cores are enabled and assigned to the logical domains in Figure K-3.
System Operation and Action
The CPU Activation Interim Permit expires 30 days later regardless of whether additional CPU Activation keys are registered in the system. In addition, a CPU Activation Interim Permit may be disabled inadvertently by the setinterimpermit command of the XSCF. Once the CPU Activation Interim Permit is disabled, it cannot be enabled again with XCP 232x. Even with XCP 2330 and later, the function cannot be enabled again instantly then and there.
If the CPU Activation Interim Permit expires or is disabled when any CPU cores are in the use violation state "VIOLATION," the system operates in the following way.
  1. When 30 minutes elapse after the violation, a notification is sent to the system, which tries to delete excessive CPU cores from logical domains.
  1. When 60 minutes elapse after the violation, the system tries to shut down the control domain.
  1. When 90 minutes elapse after the violation, the system tries to shut down all domains.
If you perform either of the following operations within 30 minutes after the violation is detected, the system does not perform operations a. to c. above.
  1. - Reducing the number of CPU cores currently used in logical domains
  1. - Register additional purchased CPU Activation keys by using the addcodactivation command of the XSCF, and then adding the number of CPU Activations to the physical partition by using the setcod command.
If the CPU Activation Interim Permit expires and you execute the showinterimpermit command of the XSCF, "expired" will be displayed as shown in the following example.
XSCF> showinterimpermit -p 0
Interim permit for PPAR 0: expired
If the CPU Activation Interim Permit expires or is disabled and you execute the showcodusage command of the XSCF, the usage of CPU cores will be displayed as shown in the following example.
XSCF> showcodusage -p resource
Resource In Use Installed CoD Permitted Status
-------- ------ --------- ------------- ------

PROC 10 16 4 VIOLATION: 6 cores in excess
In this case, "VIOLATION" is displayed in the "Status" field because the CPU Activation Interim Permit has expired. The example shows that six CPU cores are used in violation, which actually warns the system of urgency. In the case of this example, you must immediately delete six CPU cores in total from logical domains.
After detecting a violation, the system tries to delete CPU cores from logical domains until the number of CPU cores in use matches the number of purchased and registered CPU Activation keys. Six CPU cores will be deleted in this example.
The system deletes CPU cores starting with the one with the highest CPUID (CID) value. The system may not be able to delete CPU cores for some reasons, such as that a vcpu belonging to the CPU core is associated with a specific process by pbind. In that case, the CPU core with the second highest CID value will be deleted.
You can check the CID value by using the ldm command from Oracle Solaris, as shown below.
# ldm list-domain -o core
NAME
primary
CORE
CID CPUSET
0 (0, 1)
4 (8, 9)
--------------------------------------------------------------

NAME
dm0
CORE
CID CPUSET
8 (16, 17)
12 (24, 25)
<Omitted>
After the number of CPU cores becomes equal to or less than the number of CPU cores whose use is permitted by purchased CPU Activation keys, the above action is not performed.
As mentioned above, deletion of CPU cores may fail. A failure of the deletion occurs when a vcpu is associated with a specific process by pbind or other means or when the CPU core to be deleted is the last one in a logical domain. If the system cannot delete enough CPU cores (six CPU cores in this case), the control domain will be shut down.
To recover from the above situation, perform either of the following operations.
  1. Delete enough CPU cores from logical domains by using the ldm command, or stop one or more logical domains.
  1. Register a sufficient number of purchased CPU Activation keys by using the addcodactivation command of the XSCF, and then add the number of CPU Activations to the physical partition by using the setcod command.
If enough CPU cores cannot be deleted due to shut down of the control domain, all logical domains will be shut down.
If all logical domains are shut down, possible recovery methods are as follows.
  1. Select the configuration information file of the logical domains that is saved before the CPU Activation Interim Permit is enabled.
  1. Select the old configuration information file of the logical domains by using the setdomainconfig command of the XSCF, as shown below. Then, boot the system.
XSCF> setdomainconfig -p 0
PPAR-ID :0
Booting config
(Current) :IPermit-enabled
(Next) :IPermit-enabled
--------------------------------------------------------------

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

Index :2
config_name :before-IPermit
domains :2
date_created:"2015-03-24 19:21:30"
--------------------------------------------------------------

Index :3
config_name :IPermit-enabled
domains :2
date_created:"2016-06-16 12:00:30"
Select Index of Using config_name : 2
PPAR-ID of PPARs that will be affected :00
Logical domain config_name will be set to "before-IPermit".
Continue?[y|n] :y
  1. If your purpose is to recover the control domain, start only the control domain, and then delete the necessary amount of CPU core resources from the control domain.
Note - Even if the number of CPU cores used in the control domain is greater than the number of CPU cores permitted for use by purchased CPU Activation keys, the control domain is always booted normally. The same applies even if the CPU Activation Interim Permit has expired or is disabled. However, if enough CPU cores cannot be deleted within 30 minutes after the detection of the violation, the system starts automatic deletion of CPU cores again according to the above procedure. After successful deletion of CPU core resources, save the corrected logical domain configuration information by using the ldm add-spconfig command so that it can be used in the future when the control domain is rebooted.
  1. If your purpose is to recover a specific guest domain, perform the following procedure.
  1. 1. Boot only the control domain.
  1. 2. Delete CPU cores assigned to the guest domain from the control domain, and save the corrected logical domain configuration information by using the ldm add-spconfig command.
  1. 3. Boot the guest domain.
  1. Register additional purchased CPU Activation keys by using the addcodactiovation command of the XSCF, and add the number of CPU Activations to the physical partition by using the setcod command.
If you need to recover both the control domain and guest domain without using the old configuration information file of logical domains, both of the above operations b. and c. may be required.

Top of Page