Skip Navigation Links | |
Exit Print View | |
Oracle VM Server for SPARC 2.1 Administration Guide Oracle VM Server for SPARC |
Part I Oracle VM Server for SPARC 2.1 Software
1. Overview of the Oracle VM Server for SPARC Software
2. Installing and Enabling Software
4. Setting Up Services and the Control Domain
Enabling the Whole-Core Constraint
Disabling the Whole-Core Constraint
Allocating CPUs to the Control Domain
Interactions Between the Whole-Core Constraint and Other Domain Features
Using Memory Dynamic Reconfiguration
Tracking the Progress of a Memory DR Request
Memory Reconfiguration of the Control Domain
Decrease the Control Domain's Memory
Dynamic and Delayed Reconfiguration
Memory Alignment for Active Domains
Memory Alignment for Bound Domains
Memory Alignment for Inactive Domains
Using Dynamic Resource Management
Show Syntax Usage for ldm Subcommands
Utilization Statistic Definition
Generate an Extended List (-e)
Generate a Parseable, Machine-Readable List (-p)
Generate a Subset of a Long List (-o format)
List Constraints for One Domain
List Constraints in XML Format
List Constraints in a Machine-Readable Format
12. Performing Other Administration Tasks
Part II Optional Oracle VM Server for SPARC Software
13. Oracle VM Server for SPARC Physical-to-Virtual Conversion Tool
14. Oracle VM Server for SPARC Configuration Assistant
15. Using the Oracle VM Server for SPARC Management Information Base Software
16. Logical Domains Manager Discovery
17. Using the XML Interface With the Logical Domains Manager
To use Power Management (PM), you first need to set the PM mode in the Oracle Integrated Lights Out Manager (ILOM) 3.0 firmware. This section summarizes the information that you need to be able to use PM with the Oracle VM Server for SPARC software.
For more information about ILOM, see the following:
“Monitoring Power Consumption” in the Sun Integrated Lights Out Manager (ILOM) 3.0 CLI Procedures Guide
Oracle Integrated Lights Out Manager (ILOM) 3.0 Feature Updates and Release Notes
The power mode is the setting that governs system power usage at any point in time. The following power modes are supported, assuming that the underlying platform has implemented PM features:
Performance mode. The system is allowed to use all the power that is available.
Elastic mode. The system power usage is adapted to the current utilization level. For example, the power state of resources is reduced as utilization decreases.
Following are the PM features:
CPU core auto-disabling. PM automatically disables a CPU core when all the strands on that core have been disabled.
CPU clock cycle skip. Starting with the Oracle VM Server for SPARC 2.0 release, PM can automatically adjust the CPU clock cycle skip on SPARC T3 platforms. The adjustment can increase or decrease the number of clock cycles that are skipped to keep all domains within the power utilization thresholds. PM determines whether to make such adjustments based on CPU utilization. When the system goes in to performance mode, the number of clock cycles that are skipped is automatically adjusted to being none.
Memory operations in deep idle mode. Starting with the Oracle VM Server for SPARC 2.0 release, when SPARC T3 platforms are in elastic mode, they automatically configure under-utilized memory to operate in a deeper idle mode to save power.
Power limit. Starting with the Oracle VM Server for SPARC 2.1 release, you can set a power limit on SPARC T3 platforms to restrict the power draw of a system. If the power draw is greater than the power limit, PM techniques are used to reduce power. You can use the ILOM service processor (SP) to set the power limit.
See the following documents:
Sun Integrated Lights Out Manager (ILOM) 3.0 CLI Procedures Guide
Oracle Integrated Lights Out Manager (ILOM) 3.0 Feature Updates and Release Notes
You can use the ILOM interface to set a power limit, grace period, and violation action. If the power limit is exceeded for more than the grace period, the violation action is performed.
If the current power draw exceeds the power limit, an attempt is made to reduce the power state of resources that can be power-managed. If the power draw drops below the power limit, the power state of those resources is permitted to increase. If the system is in elastic mode, an increase in the power state of resources is driven by the utilization level.
When a system is in elastic mode, some domain configuration modifications are first validated to confirm that the power limit is not exceeded. If the power limit is exceeded, only some of the resources might be modified or added as requested. If the power limit is later increased, you can then add any resources that were unsuccessfully modified.
If a domain's load causes resources to consume more power, only the resources that maintain the power draw being under the power limit are successfully powered up.
For instructions on configuring the power mode by using the ILOM 3.0 firmware CLI, see “Monitoring Power Consumption” in the Sun Integrated Lights Out Manager (ILOM) 3.0 CLI Procedures Guide.
This section shows how to list power-managed strands and virtual CPUs.
# ldm list -l primary NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME primary active -n-cv- UART 64 16G 1.0% 21h 33m SOFTSTATE Solaris running UUID b9288150-327f-44f7-8c64-d4d57b92e524 MAC 00:21:28:8f:8f:34 HOSTID 0x858f8f34 CONTROL failure-policy=ignore DEPENDENCY master= CORE CID CPUSET 0 (0, 1, 2, 3, 4, 5, 6, 7) 1 (8, 9, 10, 11, 12, 13, 14, 15) 2 (16, 17, 18, 19, 20, 21, 22, 23) 3 (24, 25, 26, 27, 28, 29, 30, 31) 4 (32, 33, 34, 35, 36, 37, 38, 39) 5 (40, 41, 42, 43, 44, 45, 46, 47) 6 (48, 49, 50, 51, 52, 53, 54, 55) 7 (56, 57, 58, 59, 60, 61, 62, 63) VCPU VID PID CID UTIL STRAND 0 0 0 1.2% 100% 1 1 0 1.5% 100% 2 2 0 0.1% 100% 3 3 0 0.2% 100% . . .
The following output shows dashes (----) in the UTIL column for the CPU, which means that the strand is power-managed. The dashes are shown only for domains other than the primary domain.
# ldm list -l ldg1 NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME ldg1 active -n--v- 5000 64 16G 1.1% 20h 55m SOFTSTATE Solaris running UUID 98d86371-24f6-4792-c631-eb14e81ad4a0 MAC 00:14:4f:f9:02:f2 HOSTID 0x84f902f2 CONTROL failure-policy=ignore DEPENDENCY master= CORE CID CPUSET 8 (64, 65, 66, 67, 68, 69, 70, 71) 9 (72, 73, 74, 75, 76, 77, 78, 79) 10 (80, 81, 82, 83, 84, 85, 86, 87) 11 (88, 89, 90, 91, 92, 93, 94, 95) 12 (96, 97, 98, 99, 100, 101, 102, 103) 13 (104, 105, 106, 107, 108, 109, 110, 111) 14 (112, 113, 114, 115, 116, 117, 118, 119) 15 (120, 121, 122, 123, 124, 125, 126, 127) VCPU VID PID CID UTIL STRAND 0 64 8 0.8% 100% 1 65 8 2.0% 100% 2 66 8 ---- 100% 3 67 8 ---- 100% 4 68 8 ---- 100% . . .
In the output, a blank after util= means that the strand is power-managed.
# ldm list -l -p VCPU |vid=0|pid=0|util=0.7%|strand=100 |vid=1|pid=1|util=|strand=100 |vid=2|pid=2|util=|strand=100 |vid=3|pid=3|util=|strand=100 |vid=4|pid=4|util=0.7%|strand=100 |vid=5|pid=5|util=|strand=100 |vid=6|pid=6|util=|strand=100 |vid=7|pid=7|util=|strand=100
In the PM column of the output, a yes means that the CPU is power-managed, and a no means that the CPU is powered on. It is assumed that 100 percent free CPUs are power-managed by default, hence the dashes (---) under the PM column for them.
# ldm list-devices -a cpu VCPU PID %FREE PM 0 0 no 1 0 yes 2 0 yes 3 0 yes 4 100 --- 5 100 --- 6 100 --- 7 100 ---
In the pm= field in the output, a yes means that the CPU is power-managed, and a no means that the CPU is powered on. It is assumed that 100 percent free CPUs are power-managed by default, hence the blank in that field for them.
# ldm list-devices -a -p cpu VERSION 1.4 VCPU |pid=0|free=0|pm=no |pid=1|free=0|pm=yes |pid=2|free=0|pm=yes |pid=3|free=0|pm=yes |pid=4|free=0|pm=no |pid=5|free=0|pm=yes |pid=6|free=0|pm=yes |pid=7|free=0|pm=yes |pid=8|free=100|pm= |pid=9|free=100|pm= |pid=10|free=100|pm=