| Skip Navigation Links | |
| Exit Print View | |
|
Oracle VM Server for SPARC 2.2 Reference Manual Oracle VM Server for SPARC |
| Skip Navigation Links | |
| Exit Print View | |
|
|
Oracle VM Server for SPARC 2.2 Reference Manual Oracle VM Server for SPARC |
- command-line interface for the Logical Domains Manager
ldm or ldm --help [subcommand]
ldm -V
ldm add-domain -i file
ldm add-domain [cpu-arch=generic|native] [mac-addr=num] [hostid=num]
[failure-policy=ignore|panic|reset|stop] [extended-mapin-space=on]
[master=master-ldom1,...,master-ldom4] [max-cores=[num|unlimited]]
[uuid=uuid] [threading=max-ipc] ldom
ldm add-domain ldom...
ldm set-domain -i file
ldm set-domain [cpu-arch=generic|native] [mac-addr=num] [hostid=num]
[failure-policy=ignore|panic|reset|stop] [extended-mapin-space=[on|off]]
[master=[master-ldom1,...,master-ldom4]] [max-cores=[num|unlimited]]
[threading=[max-throughput|max-ipc]] ldom
ldm remove-domain -a
ldm remove-domain ldom...
ldm list-domain [-e] [-l] [-o format] [-p] [ldom...]
ldm migrate-domain [-f] [-n] [-p filename] source-ldom [user@]target-host[:target-ldom]
ldm add-vcpu [-c] number ldom
ldm set-vcpu [-c] number ldom
ldm remove-vcpu [-c] number ldom
ldm add-core num ldom
ldm add-core cid=core-ID[,core-ID[,...]] ldom
ldm set-core num ldom
ldm set-core cid=[core-ID[,core-ID[,...]]] ldom
ldm remove-core [-f] num ldom
ldm remove-core cid=core-ID[,core-ID[,...]] ldom
ldm add-crypto number ldom
ldm set-crypto [-f] number ldom
ldm remove-crypto [-f] number ldom
ldm add-memory [--auto-adj] size[unit] ldom
ldm add-memory mblock=PA-start:size[,PA-start:size[,...]] ldom
ldm set-memory [--auto-adj] size[unit] ldom
ldm set-memory mblock=[PA-start:size[,PA-start:size[,...]]] ldom
ldm remove-memory [--auto-adj] size[unit] ldom
ldm remove-memory mblock=PA-start:size[,PA-start:size[,...]] ldom
ldm start-reconf ldom
ldm cancel-reconf ldom
ldm cancel-operation (migration | reconf | memdr) ldom
ldm add-io (bus | device | vf-name) ldom
ldm set-io name=value [name=value ...] pf-name
ldm set-io [mac-addr=num] [alt-mac-addrs=[auto|num1,[auto|num2,...]]] [pvid=[pvid]]
[vid=[vid1,vid2,...]] [mtu=size] [name=value...] vf-name
ldm remove-io (bus | device | vf-name) ldom
ldm list-io [-l] [-p] [pf-name]
ldm list-io -d pf-name
ldm add-vsw [-q] [default-vlan-id=vlan-id] [pvid=port-vlan-id] [vid=vlan-id1,vlan-id2,...]
[linkprop=phys-state] [mac-addr=num] [net-dev=device] [mode=sc] [mtu=size]
[id=switch-id] [inter-vnet-link=on|off] vswitch-name ldom
ldm set-vsw [-q] [pvid=port-vlan-id] [vid=vlan-id1,vlan-id2,...] [mac-addr=num]
[net-dev=device] [linkprop=[phys-state]] [mode=[sc]] [mtu=size]
[inter-vnet-link=[on|off]] vswitch-name
ldm remove-vsw [-f] vswitch-name
ldm add-vnet [mac-addr=num] [mode=hybrid] [pvid=port-vlan-id] [vid=vlan-id1,vlan-id2,...]
[linkprop=phys-state] [id=network-id] [mtu=size] if-name vswitch-name ldom
ldm set-vnet [mac-addr=num] [vswitch=vswitch-name] [mode=[hybrid]] [pvid=port-vlan-id]
[linkprop=[phys-state]] [vid=vlan-id1,vlan-id2,...] [mtu=size] if-name ldom
ldm remove-vnet [-f] if-name ldom
ldm add-vds service-name ldom
ldm remove-vds [-f] service-name
ldm add-vdsdev [-f] [-q] [options={ro,slice,excl}] [mpgroup=mpgroup] backend
volume-name@service-name
ldm set-vdsdev [-f] options=[{ro,slice,excl}] [mpgroup=mpgroup]
volume-name@service-name
ldm remove-vdsdev [-f] volume-name@service-name
ldm add-vdisk [timeout=seconds] [id=disk-id] disk-name volume-name@service-name ldom
ldm set-vdisk [timeout=seconds] [volume=volume-name@service-name] disk-name ldom
ldm remove-vdisk [-f] disk-name ldom
ldm add-vdpcs vdpcs-service-name ldom
ldm remove-vdpcs [-f] vdpcs-service-name
ldm add-vdpcc vdpcc-name vdpcs-service-name ldom
ldm remove-vdpcc [-f] vdpcc-name ldom
ldm add-vcc port-range=x-y vcc-name ldom
ldm set-vcc port-range=x-y vcc-name
ldm remove-vcc [-f] vcc-name
ldm set-vcons [port=[port-num]] [group=group] [service=vcc-server] ldom
ldm create-vf [mac-addr=num] [alt-mac-addrs=[auto|num1,[auto|num2,...]]] [pvid=pvid]
[vid=vid1,vid2,...] [mtu=size] [name=value...] pf-name
ldm destroy-vf vf-name
ldm add-variable var-name=[value]... ldom
ldm set-variable var-name=[value]... ldom
ldm remove-variable var-name... ldom
ldm list-variable [var-name...] ldom
ldm start-domain (-a | -i file | ldom...)
ldm stop-domain [-f] (-a | ldom...)
ldm panic-domain ldom
ldm bind-domain [-f] [-q] (-i file | ldom)
ldm unbind-domain ldom
ldm list-bindings [-e] [-p] [ldom...]
ldm add-spconfig config-name
ldm add-spconfig -r autosave-name [new-config-name]
ldm set-spconfig config-name
ldm set-spconfig factory-default
ldm remove-spconfig [-r] config-name
ldm list-spconfig [-r [autosave-name]]
ldm list-constraints ([-x] | [-e] [-p]) [ldom...]
ldm list-devices [-a] [-p] [core] [cpu] [crypto] [memory] [io]
ldm list-services [-e] [-p] [ldom...]
ldm add-policy [enable=yes|no] [priority=value] [attack=value] [decay=value]
[elastic-margin=value] [sample-rate=value] [tod-begin=hh:mm[:ss]]
[tod-end=hh:mm[:ss]] [util-lower=percent] [util-upper=percent] [vcpu-min=value]
[vcpu-max=value] name=policy-name ldom...
ldm set-policy [enable=[yes|no]] [priority=[value]] [attack=[value]] [decay=[value]]
[elastic-margin=[value]] [sample-rate=[value]] [tod-begin=[hh:mm:ss]]
[tod-end=[hh:mm:ss]] [util-lower=[percent]] [util-upper=[percent]] [vcpu-min=[value]]
[vcpu-max=[value]] name=policy-name ldom...
ldm remove-policy [name=]policy-name... ldom
ldm init-system [-frs] -i fileA logical domain is a discrete logical grouping with its own operating system, resources, and identity within a single computer system. Each logical domain can be created, destroyed, reconfigured, and rebooted independently, without requiring a power cycle of the server. You can use logical domains to run a variety of applications in different domains and keep them independent for security purposes.
All logical domains are the same and can be distinguished from one another based on the roles that you specify for them. The following are the roles that logical domains can perform:
Creates and manages other logical domains and services by communicating with the hypervisor.
Provides services to other logical domains, such as a virtual network switch or a virtual disk service.
Has direct access to a physical I/O device, such as a network card in a PCI EXPRESS (PCIe) controller or a single-root I/O virtualization (SR-IOV) virtual function. An I/O domain can own a PCIe root complex, or it can own a PCIe slot or on-board PCIe device by using the direct I/O (DIO) feature and an SR-IOV virtual function by using the SR-IOV feature.
An I/O domain can share physical I/O devices with other domains in the form of virtual devices when the I/O domain is also used as a service domain.
Has a PCIe root complex assigned to it. This domain owns the PCIe fabric and all connected devices, and provides all fabric-related services, such as fabric error handling. A root domain owns all of the SR-IOV physical functions from which you can create virtual functions and assign them to I/O domains. A root domain is also an I/O domain, as it owns and has direct access to physical I/O devices.
The number of root domains that you can have depends on your platform architecture. For example, if you are using a Sun SPARC Enterprise T5440 server from Oracle, you can have up to four root domains.
Uses services from the I/O and service domains and is managed by the control domain.
You can use the Logical Domains Manager to establish dependency relationships between domains.
A domain that has one or more domains that depend on it. A master domain specifies a failure policy to be enacted by its slave domains when the master domain fails. For instance, depending on the master domain's failure policy, a slave can be left as-is, panicked, rebooted, or stopped when the master domain fails.
A domain that depends on another domain. A domain can specify up to four master domains that dictate the failure policy to enact when one or more of the master domains fail.
Following are the supported subcommands along with a description and required authorization for each. For information about setting up authorization for user accounts, see Creating Authorizations and Profiles and Assigning Roles to User Accounts in Oracle VM Server for SPARC 2.2 Administration Guide.
|
Note - Not all subcommands are supported on all resources types.
The following table shows the three kinds of aliases for ldm subcommands.
|
Note - In the syntax and examples in the remainder of this man page, the short forms of the action and resource aliases are used.
The following resources are supported:
CPU cores.
Any supported cryptographic unit on a supported server. Currently, the two cryptographic units supported are the Modular Arithmetic Unit (MAU) and the Control Word Queue (CWQ).
I/O devices, such as PCIe root complexes and their attached adapters and devices. Also direct I/O-assignable devices and PCIe SR-IOV virtual functions.
Default memory size in bytes. Or specify gigabytes (G), kilobytes (K), or megabytes (M). Virtualized memory of the server that can be allocated to guest domains.
Virtual console concentrator service with a specific range of TCP ports to assign to each guest domain at the time it is created.
Virtual console for accessing system-level messages. A connection is achieved by connecting to the vconscon service in the control domain at a specific port.
Each virtual CPU represents one CPU thread of a server. For example, an 8-core Sun SPARC Enterprise T5120 server has 64 CPU threads (virtual CPUs) that can be allocated among the logical domains.
Virtual disks are generic block devices backed by different types of physical devices, volumes, or files. A virtual disk is not synonymous with a SCSI disk and, therefore, excludes the target ID (tN) in the disk name. Virtual disks in a logical domain have the following format: cNdNsN, where cN is the virtual controller, dN is the virtual disk number, and sN is the slice.
Virtual disk server that allows you to export virtual disks to other logical domains.
Device exported by the virtual disk server. The device can be an entire disk, a slice on a disk, a file, or a disk volume.
Virtual data plane channel client. Only of interest in a Netra Data Plane Software (NDPS) environment.
Virtual data plane channel service. Only of interest in a Netra Data Plane Software (NDPS) environment.
Virtual network device that implements a virtual Ethernet device and communicates with other vnet devices in the system using the virtual network switch (vsw).
Virtual network switch that connects the virtual network devices to the external network and also switches packets between them.
The following list types are supported:
Lists the resources bound to a logical domain.
Lists the logical domain configurations stored on the service processor (SP). On SPARC T3 based servers, the configurations are shown in the order of creation from the oldest to the newest.
Lists the constraints used to create a logical domain.
Lists all free devices for the server.
Lists all services exported by a logical domain.
The following table describes the ldm command options. The short form of the option is followed by the long form, if applicable.
|
Note - You can use various ldm set-* commands to reset any property to its default value by specifying an empty value. For example, the following ldm set-policy command resets the attack property to its default value:
# ldm set-policy attack= high-policy ldom1
The following properties are supported:
Specifies a comma-separated list of alternate MAC addresses. Valid values are numeric MAC addresses and the auto keyword, which can be used one or more times to request that the system generate an alternate MAC address. The auto keyword can be mixed with numeric MAC addresses.
Specifies the maximum number of resources to be added during any one resource control cycle. If the number of available resources is less than the specified value, all of the available resources are added. By default, the attack is unlimited so that you can add as many CPU threads as are available. Valid values are from 1 to the number of free CPU threads on the system.
Specifies the physical core IDs to assign to or remove from a domain. To remove all named cores, omit core-ID values for the cid property by running the ldm set-core cid= command.
The cid property should only be used by an administrator who is knowledgeable about the topology of the system to be configured. This advanced configuration feature enforces specific allocation rules and might affect the overall performance of the system.
Specifies one of the following values:
generic uses common CPU hardware features to enable a guest domain to perform a CPU-type-independent migration.
native uses CPU-specific hardware features to enable a guest domain to migrate only between platforms that have the same CPU type. native is the default value.
Using the generic value might result in reduced performance compared to the native value. This occurs because the guest domain does not use some features that are only present in newer CPU types. By not using these features, the generic setting enables the flexibility of migrating the domain between systems that use newer and older CPU types.
Specifies the maximum number of resources to be removed during any one resource control cycle. Only the number of currently bound CPU threads minus the value of vcpu-min can be removed even if the value specified by this property is larger. By default, the value is 1. Valid values are from 1 to the total number of CPU threads minus 1.
Specifies the default virtual local area network (VLAN) to which a virtual network device or virtual switch needs to be a member, in tagged mode. The first VLAN ID (vid1) is reserved for the default-vlan-id.
Specifies the amount of buffer space between util-lower and the number of free CPU threads to avoid oscillations at low CPU thread counts. Valid values are from 0 to 100. The default value is 5.
Enables or disables resource management for an individual domain. By default, enable=yes.
Enables or disables extended mapin space for a domain. By default, extended-mapin-space=off.
The extended mapin space refers to the additional LDC shared memory space. This memory space is required to support a large number of virtual I/O devices that use direct-mapped shared memory. This extended mapin space is also used by virtual network devices to improve performance and scalability.
Specifies the master domain's failure policy, which controls how slave domains behave when the master domain fails. This property is set on a master domain. The default value is ignore. Following are the valid property values:
ignore ignores failures of the master domain (slave domains are unaffected).
panic panics any slave domains when the master domain fails.
reset resets any slave domains when the master domain fails.
stop stops any slave domains when the master domain fails.
Specifies a group to which to attach a console. The group argument allows multiple consoles to be multiplexed onto the same TCP connection.
Specifies the host ID for a particular domain. If you do not specify a host ID, the Logical Domains Manager assigns a unique host ID to each domain.
Specifies an ID for a new virtual disk device, virtual network device, and virtual switch device, respectively.
Specifies whether to assign a channel between each virtual network device. The default value is on.
When inter-vnet-link=on, the Logical Domains Manager assigns a channel between each pair of virtual network devices that are connected to the same virtual switch for better guest-to-guest performance.
When inter-vnet-link=off, the Logical Domains Manager only assigns channels for communications between virtual network devices and virtual switches. In this case, guest-to-guest communications traffic goes through the virtual switch. This setting reduces the number of channels that are used for virtual network devices. Thus, the maximum number of virtual devices that you can add to the system is increased.
Specifies whether the virtual device reports its link status based on the underlying physical network device. When linkprop=phys-state is specified on the command line, the virtual device link status reflects the physical link state. By default, the virtual device link status does not reflect the physical link state.
Defines a MAC address. The number must be in standard octet notation, for example, 80:00:33:55:22:66.
Specifies the name of up to four master domains for a slave domain. This property is set on a slave domain. By default, there are no masters for the domain. The domain must already exist prior to an ldm add-domain operation.
Note - The Logical Domains Manager does not permit you to create domain relationships that result in a dependency cycle.
Specifies the maximum number of cores that are permitted to be assigned to a domain. If the value is unlimited, there is no constraint on the number of CPU cores that can be allocated.
Specifies one or more physical memory blocks to assign to or remove from a domain. PA-start specifies the starting physical address of the memory block in hexadecimal format. size is the size of the memory block, including a unit, to be assigned to or removed from the domain. To remove all named memory blocks, omit PA-start:size values from the mblock property by running the ldm set-memory mblock= command.
The mblock property should only be used by an administrator who is knowledgeable about the topology of the system to be configured. This advanced configuration feature enforces specific allocation rules and might affect the overall performance of the system.
For add-vsw and set-vsw subcommands:
Omit this option when you are not running Oracle Solaris Cluster software in guest domains because you could impact virtual network performance.
Otherwise, specify one of the following:
Set mode=sc to enable virtual networking support for prioritized processing of Oracle Solaris Cluster heartbeat packets in a Logical Domains environment.
Leave the mode= argument blank in the set-vsw subcommand to stop special processing of heartbeat packets.
For add-vnet and set-vnet subcommands:
Omit this option when you do not want to use NIU Hybrid I/O.
Otherwise, specify one of the following:
Set mode=hybrid to request the system to use NIU Hybrid I/O if possible. If it is not possible, the system reverts to virtual I/O. See Using NIU Hybrid I/O in Oracle VM Server for SPARC 2.2 Administration Guide.
Leave the mode= argument blank in the set-vnet subcommand to disable NIU Hybrid I/O.
Defines the multipath group name for several virtual disk server devices (vdsdev). So, when a virtual disk cannot communicate with a virtual disk server device, a failover is initiated to another virtual disk server device in the multipath group.
Specifies the maximum transmission unit (MTU) of a virtual switch, virtual network devices that are bound to the virtual switch, or both. Valid values are in the range of 1500-16000. The ldm command issues an error if an invalid value is specified.
Specifies the resource management policy name.
Defines the path name of the actual network device.
Specifies all or a subset of the following options for a specific virtual disk server device. Separate two or more options with commas and no spaces, such as ro,slice,excl.
ro – Specifies read-only access
slice – Exports a back end as a single slice disk
excl – Specifies exclusive disk access
Omit the options= argument or leave it blank in an add-vdsdev subcommand to have the default values of disk, not exclusive, and read/write. Leave the options= argument blank in the set-vdsdev subcommand to turn off any previous options specified.
Specifies a specific port number or, left blank, lets the Logical Domains Manager set the port number.
Defines a range of TCP ports.
Specifies a priority for dynamic resource management (DRM) policies. Priority values are used to determine the relationship between DRM policies in a single domain and between DRM-enabled domains in a single system. Lower numerical values represent higher (better) priorities. Valid values are between 1 and 9999. The default value is 99.
The behavior of the priority property depends on whether a pool of free CPU resources is available, as follows:
Free CPU resources are available in the pool. In this case, the priority property determines which DRM policy will be in effect when more than one overlapping policy is defined for a single domain.
No free CPU resources are available in the pool. In this case, the priority property specifies whether a resource can be dynamically moved from a lower-priority domain to a higher-priority domain in the same system. The priority of a domain is the priority specified by the DRM policy that is in effect for that domain.
For example, a higher-priority domain can acquire CPU resources from another domain that has a DRM policy with a lower priority. This resource-acquisition capability pertains only to domains that have DRM policies enabled. Domains that have equal priority values are unaffected by this capability. So, if the default priority is used for all policies, domains cannot obtain resources from lower-priority domains. To take advantage of this capability, adjust the priority property values so that they have unequal values.
Specifies the VLAN to which the virtual network device needs to be a member, in untagged mode.
Specifies the cycle time, in seconds, which is the sample rate for DRM. Valid values are from 1 to 9999. The default and recommended value is 10.
Specifies the name of the existing virtual console concentrator that you want to handle the console connection.
The threading property specifies the workflow throughput of the domain.
The following are valid values for the threading property:
max-ipc. Only one thread is active for each CPU core that is assigned to the domain, which maximizes the number of instructions per cycle. Selecting this mode requires that the domain is also configured with the whole-core constraint. See the add-vcpu and set-vcpu subcommand descriptions.
max-throughput. Activates all threads that are assigned to the domain, which maximizes throughput. This mode is used by default and is also selected if you do not specify any mode (threading=).
Defines the number of seconds for establishing a connection between a virtual disk client (vdc) and a virtual disk server (vds). If there are multiple virtual disk (vdisk) paths, then the vdc can try to connect to a different vds, and the timeout ensures that a connection to any vds is established within the specified amount of time. Specify 0 to disable the timeout in the set-vdisk subcommand.
Specifies the effective start time of a policy in terms of hour, minute, and optional second. This time must be earlier than the time specified by tod-end in a period that begins at midnight and ends at 23:59:59. The default value is 00:00:00.
Specifies the effective stop time of a policy in terms of hour, minute, and optional second. This time must be later than the time specified by tod-begin in a period that begins at midnight and ends at 23:59:59. The default value is 23:59:59.
Specifies the lower utilization level at which policy analysis is triggered. Valid values are from 1 to util-upper minus 1. The default value is 60.
Specifies the upper utilization level at which policy analysis is triggered. Valid values are from util-lower plus 1 to 99. The default value is 85.
Specifies the universally unique identifier (UUID) for the domain. uuid is a hexadecimal string, such as 12345678-1234-abcd-1234-123456789abc, which consists of five hexadecimal numbers separated by dashes. Each number must have the specified number of hexadecimal digits: 8, 4, 4, 4, and 12, as follows:
xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
Specifies the maximum number of CPU thread resources for a domain. By default, the maximum number of CPU threads is unlimited. Valid values are from vcpu-min plus 1 to the total number of free CPU threads on the system.
Specifies the minimum number of CPU thread resources for a domain. Valid values are from 1 to vcpu-max minus 1. The default value is 1.
Specifies the VLAN to which a virtual network device or virtual switch needs to be a member, in tagged mode.
Changes a volume name for a virtual disk.
Changes a virtual switch name for a virtual network.
Following are definitions of the flags in the list subcommand output:
Placeholder
Control domain
Delayed reconfiguration
Error
Normal
Memory DR in progress
Column 1 – starting or stopping
Column 6 – source domain
Column 2 – transition
Column 6 – target domain
Virtual I/O service domain
The list flag values are position dependent. Following are the values that can appear in each of the five columns from left to right.
Table 1 List Flag Positions
|
This section contains descriptions of every supported command-line interface (CLI) operation, that is, every subcommand and resource combination.
This subcommand adds one or more logical domains by specifying one or more logical domain names or by using an XML configuration file. You can also specify property values to customize the domain, such as the MAC address, the host ID, a list of master domains, and a failure policy. If you do not specify these property values, the Logical Domains Manager automatically assigns default values.
ldm add-dom -i file ldm add-dom [cpu-arch=generic|native] [mac-addr=num] [hostid=num] [failure-policy=ignore|panic|reset|stop] [extended-mapin-space=on] [master=master-ldom1,...,master-ldom4] [max-cores=[num|unlimited]] [uuid=uuid] [threading=max-ipc] ldom ldm add-dom ldom...
where:
-i file specifies the XML configuration file to use in creating the logical domain.
cpu-arch=generic|native specifies one of the following values:
generic configures a guest domain for a CPU-type-independent migration.
native configures a guest domain to migrate only between platforms that have the same CPU type. native is the default value.
mac-addr=num is the MAC address for this domain. The number must be in standard octet notation, for example, 80:00:33:55:22:66.
hostid specifies the host ID for a particular domain. If you do not specify a host ID, the Logical Domains Manager assigns a unique host ID to each domain.
failure-policy specifies the master domain's failure policy, which controls how slave domains behave when the master domain fails. This property is set on a master domain. The default value is ignore. Following are the valid property values:
ignore ignores failures of the master domain (slave domains are unaffected).
panic panics any slave domains when the master domain fails.
reset resets any slave domains when the master domain fails.
stop stops any slave domains when the master domain fails.
extended-mapin-space=on enables the extended mapin space for the specified domain. By default, the extended mapin space is disabled.
master specifies the name of up to four master domains for a slave domain. This property is set on a slave domain. By default, there are no masters for the domain. The master domain must exist prior to an ldm add-domain operation.
Note - The Logical Domains Manager does not permit you to create domain relationships that result in a dependency cycle.
Setting the threading property specifies the workflow throughput of the domain.
The following are valid values for the threading property:
max-ipc. Only one thread is active for each CPU core that is assigned to the domain, which maximizes the number of instructions per cycle. Selecting this mode requires that the domain is also configured with the whole-core constraint. See the add-vcpu and set-vcpu subcommand descriptions.
max-throughput. Activates all threads that are assigned to the domain, which maximizes throughput. This mode is used by default and is also selected if you do not specify any mode (threading=).
uuid=uuid specifies the universally unique identifier (UUID) for the domain. uuid is a hexadecimal string, such as 12345678-1234-abcd-1234-123456789abc, which consists of five hexadecimal numbers separated by dashes. Each number must have the specified number of hexadecimal digits: 8, 4, 4, 4, and 12, as follows:
xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
max-cores=[num|unlimited] specifies the maximum number of cores that are permitted to be assigned to a domain. If the value is unlimited, there is no constraint on the number of CPU cores that can be allocated.
ldom specifies the logical domain to be added.
This subcommand enables you to modify only the mac-addr, hostid, failure-policy, extended-mapin-space, master, max-cores, and threading properties of each domain. You cannot use this command to update resource properties.
Note - If the slave domain is bound, all of its specified master domains must also be bound prior to invoking the ldm set-domain command.
ldm set-dom -i file ldm set-dom [cpu-arch=generic|native] [mac-addr=num] [hostid=num] [failure-policy=ignore|panic|reset|stop] [extended-mapin-space=[on|off]] [master=[master-ldom1,...,master-ldom4]] [max-cores=[num|unlimited]] [threading=[max-throughput|max-ipc]] ldom
where:
-i file specifies the XML configuration file to use in creating the logical domain.
Only the ldom_info nodes specified in the XML file are parsed. Resource nodes, such as vcpu, mau, and memory, are ignored.
cpu-arch=generic|native specifies one of the following values:
generic configures a guest domain for a CPU-type-independent migration.
native configures a guest domain to migrate only between platforms that have the same CPU type. native is the default value.
mac-addr=num is the MAC address for this domain. The number must be in standard octet notation, for example, 80:00:33:55:22:66.
hostid specifies the host ID for a particular domain. If you do not specify a host ID, the Logical Domains Manager assigns a unique host ID to each domain.
failure-policy specifies the master domain's failure policy, which controls how slave domains behave when the master domain fails. This property is set on a master domain. The default value is ignore. Following are the valid property values:
ignore ignores failures of the master domain (slave domains are unaffected).
panic panics any slave domains when the master domain fails.
reset resets any slave domains when the master domain fails.
stop stops any slave domains when the master domain fails.
extended-mapin-space enables or disables the extended mapin space for the specified domain. By default, the extended-mapin-space=off, which is equivalent to setting extended-mapin-space=.
master specifies the name of up to four master domains for a slave domain. This property is set on a slave domain. By default, there are no masters for the domain. The master domain must already exist prior to this operation.
Note - The Logical Domains Manager does not permit you to create domain relationships that result in a dependency cycle.
Setting the threading property specifies the workflow throughput of the domain.
The following are valid values for the threading property:
max-ipc. Only one thread is active for each CPU core that is assigned to the domain, which maximizes the number of instructions per cycle. Selecting this mode requires that the domain is also configured with the whole-core constraint. See the add-vcpu and set-vcpu subcommand descriptions.
max-throughput. Activates all threads that are assigned to the domain, which maximizes throughput. This mode is used by default and is also selected if you do not specify any mode (threading=).
max-cores=[num|unlimited] specifies the maximum number of cores that are permitted to be assigned to a domain. If the value is unlimited, there is no constraint on the number of CPU cores that can be allocated.
ldom specifies the name of the logical domain for which you want to set options.
This subcommand removes one or more logical domains.
ldm rm-dom -a ldm rm-dom ldom...
where:
-a deletes all logical domains except the control domain.
ldom specifies the logical domain to be deleted.
In the event that the domain to be destroyed is specified as a master domain, references to this domain are removed from all slave domains.
This subcommand migrates a domain from one location to another.
ldm migrate-domain [-f] [-n] [-p filename] source-ldom [user@]target-host[:target-ldom]
where:
-f attempts to force the migration of the domain.
-n performs a dry run on the migration to determine whether it will succeed. It does not actually migrate the domain.
-p filename reads the password needed on the target machine from the first line of filename. This option enables you to perform non-interactive migrations that do not require you to provide the target machine password at a prompt.
If you plan to store passwords in this manner, ensure that the file permissions are set so that only the root owner, or a privileged user, can read or write the file (400 or 600).
source-ldom is the logical domain that you want to migrate.
user is the user name that is authorized to run the Logical Domains Manager on the target host. If no user name is specified, the name of the user running the command is used by default.
target-host is the host where you want to place the target-ldom.
target-ldom is the logical domain name to be used on the target machine. The default is to keep the domain name used on the source domain (source-ldom).
Logical Domains supports the following types of reconfiguration operations:
Dynamic reconfiguration operations. DR is the ability to add, set, or remove resources to or from an active domain. The ability to perform dynamic reconfiguration of a particular resource type is dependent on having support in the particular version of the OS running in the logical domain. For the control domain, if a DR cannot be done, a delayed reconfiguration operation is done instead. You can also manually enter delayed reconfiguration mode on the control domain by running the ldm start-reconf primary command.
Delayed reconfiguration operations. In contrast to DR operations that take place immediately, delayed reconfiguration operations take effect after the next reboot of the OS or stop and start of the logical domain if no OS is running. Delayed reconfiguration operations can only be performed on the control domain. Other domains must be stopped prior to modifying resources that cannot be dynamically configured.
See Resource Reconfiguration in Oracle VM Server for SPARC 2.2 Administration Guide for more information about dynamic reconfiguration and delayed reconfiguration.
You can allocate either CPU threads or CPU cores to a domain. To allocate CPU threads, use the add-vcpu, set-vcpu, and remove-vcpu subcommands. To allocate CPU cores, use the add-core, set-core, and remove-core subcommands.
Add CPU ThreadsThis subcommand adds the specified number of CPU threads or CPU cores to a logical domain. Note that a domain cannot be configured simultaneously with CPU cores and CPU threads. CPU core configurations and CPU thread configurations are mutually exclusive.
ldm add-vcpu [-c] number ldom
where:
-c performs the following discrete CPU operations:
Sets the allocation unit for the domain from threads to cores, if not already set, and adds the specified number of cores to the domain.
If the domain is inactive, sets a cap on the number of cores that can be allocated to the domain when it is bound or active. A cap is set on the primary domain only if the domain is in a delayed reconfiguration mode.
If any allocation request results in more cores being assigned to a domain than is permitted by the cap, the command fails.
This option configures hard partitioning on your Oracle VM Server for SPARC system. See Configuring the System With Hard Partitions in Oracle VM Server for SPARC 2.2 Administration Guide.
You can change the allocation unit from cores to threads and remove the cap. Make these changes by issuing an add-vcpu, set-vcpu, or rm-vcpu command without the -c option on an inactive domain or on the primary domain that is in delayed reconfiguration mode.
Starting with the Oracle VM Server for SPARC 2.2 release, the CPU cap and the allocation of CPU cores is handled by separate commands. By using these commands you can independently allocate CPU cores, set a cap, or both. The allocation unit can be set to cores even when no cap is in place. However, running the system in this mode is not acceptable for configuring hard partitioning on your Oracle VM Server for SPARC system.
Allocate the specified number of CPU cores to a domain by using the add-core, set-core, and rm-core subcommands.
Set the cap by using the create-domain or set-domain subcommand to specify the max-cores property value.
When the -c option is not specified, number is the number of CPU threads to be added to the logical domain. When the -c option is specified, number is the number of CPU cores to be added to the logical domain.
ldom specifies the logical domain where the CPU threads are to be added.
This subcommand specifies the number of CPU threads or CPU cores to be set in a logical domain. Note that a domain cannot be configured simultaneously with CPU cores and CPU threads. CPU core configurations and CPU thread configurations are mutually exclusive.
ldm set-vcpu [-c] number ldom
where:
-c performs the following discrete CPU operations:
Sets the allocation unit for the domain from threads to cores, if not already set, and sets the allocation to the specified number of cores.
If the domain is inactive, sets a cap on the number of cores that can be allocated to the domain when it is bound or active. A cap is set on the primary domain only if the domain is in a delayed reconfiguration mode.
If any allocation request results in more cores being assigned to a domain than is permitted by the cap, the command fails.
This option configures hard partitioning on your Oracle VM Server for SPARC system. See Configuring the System With Hard Partitions in Oracle VM Server for SPARC 2.2 Administration Guide.
You can change the allocation unit from cores to threads and remove the cap. Make these changes by issuing an add-vcpu, set-vcpu, or rm-vcpu command without the -c option on an inactive domain or on the primary domain that is in delayed reconfiguration mode.
Starting with the Oracle VM Server for SPARC 2.2 release, the CPU cap and the allocation of CPU cores is handled by separate commands. By using these commands you can independently allocate CPU cores, set a cap, or both. The allocation unit can be set to cores even when no cap is in place. However, running the system in this mode is not acceptable for configuring hard partitioning on your Oracle VM Server for SPARC system.
Allocate the specified number of CPU cores to a domain by using the add-core, set-core, and rm-core subcommands.
Set the cap by using the create-domain or set-domain subcommand to specify the max-cores property value.
When the -c option is not specified, number is the number of CPU threads to be set in a logical domain. When the -c option is specified, number is the number of CPU cores to be set in a logical domain.
ldom is the logical domain where the number of CPU threads are to be set.
This subcommand removes the specified number of CPU threads or CPU cores from a logical domain. Note that a domain cannot be configured simultaneously with CPU cores and CPU threads. CPU core configurations and CPU thread configurations are mutually exclusive.
ldm rm-vcpu [-c] number ldom
where:
-c performs the following discrete CPU operations:
Sets the allocation unit for the domain from threads to cores, if not already set, and removes the specified number of cores from the domain.
If the domain is inactive, sets a cap on the number of cores that can be allocated to the domain when it is bound or active. A cap is set on the primary domain only if the domain is in a delayed reconfiguration mode.
If any allocation request results in more cores being assigned to a domain than is permitted by the cap, the command fails.
This option configures hard partitioning on your Oracle VM Server for SPARC system. See Configuring the System With Hard Partitions in Oracle VM Server for SPARC 2.2 Administration Guide.
You can change the allocation unit from cores to threads and remove the cap. Make these changes by issuing an add-vcpu, set-vcpu, or rm-vcpu command without the -c option on an inactive domain or on the primary domain that is in delayed reconfiguration mode.
Starting with the Oracle VM Server for SPARC 2.2 release, the CPU cap and the allocation of CPU cores is handled by separate commands. By using these commands you can independently allocate CPU cores, set a cap, or both. The allocation unit can be set to cores even when no cap is in place. However, running the system in this mode is not acceptable for configuring hard partitioning on your Oracle VM Server for SPARC system.
Allocate the specified number of CPU cores to a domain by using the add-core, set-core, and rm-core subcommands.
Set the cap by using the create-domain or set-domain subcommand to specify the max-cores property value.
When the -c option is not specified, number is the number of CPU threads to be removed from the logical domain. When the -c option is specified, number is the number of CPU cores to be removed from the logical domain.
ldom specifies the logical domain where the CPU threads are to be removed.
This subcommand adds the specified number of CPU cores to a domain. When you specify the number of CPU cores, the cores to be assigned are automatically selected. However, when you specify a core-ID value to the cid property, the specified cores are explicitly assigned.
The cid property should only be used by an administrator who is knowledgeable about the topology of the system to be configured. This advanced configuration feature enforces specific allocation rules and might affect the overall performance of the system.
You can explicitly assign cores to a domain as long as power management is not using the elastic policy.
ldm add-core num ldom ldm add-core cid=core-ID[,core-ID[,...]] ldom
where:
num specifies the number of CPU cores to assign to a domain.
cid=core-ID[,...] specifies one or more physical CPU cores to assign to a domain.
ldom specifies the domain to which the CPU cores are assigned.
This subcommand specifies the number of CPU cores to assign to a domain. When you specify the number of CPU cores, the cores to be assigned are automatically selected. However, when you specify a core-ID value to the cid property, the specified cores are explicitly assigned.
You can explicitly assign cores to a domain as long as power management is not using the elastic policy.
ldm set-core num ldom ldm set-core cid=[core-ID[,core-ID[,...]]] ldom
where:
num specifies the number of CPU cores to assign to a domain.
cid=core-ID[,...] specifies one or more physical CPU cores to assign to a domain. cid= removes all named CPU cores.
ldom specifies the domain to which the CPU cores are assigned.
This subcommand specifies the number of CPU cores to remove from a domain. When you specify the number of CPU cores, the cores to be removed are automatically selected. However, when you specify a core-ID value to the cid property, the specified cores are explicitly removed.
You can explicitly remove cores from a domain as long as power management is not using the elastic policy.
ldm remove-core [-f] num ldom ldm remove-core cid=[core-ID[,core-ID[,...]]] ldom
where:
-f attempts to force the removal of one or more cores from an active domain.
num specifies the number of CPU cores to remove from a domain.
cid=core-ID[,...] specifies one or more physical CPU cores to remove from a domain.
ldom specifies the domain from which the CPU cores are removed.
The cryptographic unit subcommands only pertain to SPARC platforms that have discrete cryptographic units.
Add Cryptographic UnitsThis subcommand specifies the number of cryptographic units to be added to a logical domain. Currently, the supported cryptographic units on supported servers are the Modular Arithmetic Unit (MAU) and the Control Word Queue (CWQ).
ldm add-crypto number ldom
where:
number is the number of cryptographic units to be added to the logical domain.
ldom specifies the logical domain where the cryptographic units are to be added.
This subcommand specifies the number of cryptographic units to be set in a logical domain. If you want to remove all cryptographic units from an active domain, you must specify the -f option.
To remove the last cryptographic unit from the primary domain when domains are active, you must do one of the following:
Use DR and specify the -f option
Use delayed reconfiguration
ldm set-crypto [-f] number ldom
where:
-f forces the removal of the last cryptographic unit in the domain if number is 0.
The -f option is only necessary in the following situations:
When the guest domain is active
On the primary domain, but only if at least one active guest domain exists on the system
number is the number of cryptographic units to be set in the logical domain.
ldom specifies the logical domain where the number of cryptographic units are to be set.
This subcommand removes the specified number of cryptographic units from a logical domain. If you want to remove all cryptographic units from an active domain, you must specify the -f option.
To remove the last cryptographic unit from the primary domain when domains are active, you must do one of the following:
Use DR and specify the -f option
Use delayed reconfiguration
ldm rm-crypto [-f] number ldom
where:
-f forces the removal of the last cryptographic unit in the domain if number is equal to the number of cryptographic units in the domain.
The -f option is only necessary in the following situations:
When the guest domain is active
On the primary domain, but only if at least one active guest domain exists on the system
number is the number of cryptographic units to be removed from the logical domain.
ldom specifies the logical domain where the cryptographic units are to be removed.
This subcommand adds the specified amount of memory to a domain. When you specify a memory block size, the memory block to be assigned is automatically selected. However, when you specify a PA-start:size value to the mblock property, the specified memory blocks are explicitly assigned.
The mblock property should only be used by an administrator who is knowledgeable about the topology of the system to be configured. This advanced configuration feature enforces specific allocation rules and might affect the overall performance of the system.
ldm add-mem [--auto-adj] size[unit] ldom ldm add-mem mblock=PA-start:size[,PA-start:size[,...]] ldom
where:
--auto-adj specifies that the amount of memory to be added to an active domain is automatically 256Mbyte-aligned, which might increase the requested memory size. If the domain is inactive, bound, or in a delayed reconfiguration, this option automatically aligns the resulting size of the domain by rounding up to the next 256-Mbyte boundary.
size is the size of memory to be added to a logical domain.
unit is the unit of measurement. The default is bytes. If you want a different unit of measurement, specify one of the following (the unit is not case-sensitive):
G is gigabytes
K is kilobytes
M is megabytes
mblock=PA-start:size specifies one or more physical memory blocks to assign to a domain. PA-start specifies the starting physical address of the memory block in hexadecimal format. size is the size of the memory block, including a unit, to be assigned to the domain.
ldom specifies the logical domain where the memory is to be added.
This subcommand sets a specific amount of memory in a domain. Depending on the amount of memory specified, this subcommand is treated as an add-memory or remove-memory operation.
When you specify a memory block size, the memory block to be assigned is automatically selected. However, when you specify a PA-start:size value to the mblock property, the specified memory blocks are explicitly assigned.
ldm set-mem [--auto-adj] size[unit] ldom ldm set-mem mblock=PA-start:size[,PA-start:size[,...]] ldom
where:
--auto-adj specifies that the amount of memory to be set on an active domain is automatically 256Mbyte-aligned, which might increase the requested memory size. If the domain is inactive, bound, or in a delayed reconfiguration, this option automatically aligns the resulting size of the domain by rounding up to the next 256-Mbyte boundary.
size is the size of memory to be set in the logical domain.
unit is the unit of measurement. The default is bytes. If you want a different unit of measurement, specify one of the following (the unit is not case-sensitive):
G is gigabytes
K is kilobytes
M is megabytes
mblock=PA-start:size specifies one or more physical memory blocks to assign to a domain. PA-start specifies the starting physical address of the memory block in hexadecimal format. size is the size of the memory block, including a unit, to be assigned to the domain.
ldom specifies the logical domain where the memory is to be modified.
This subcommand removes the specified amount of memory from a logical domain. When you specify a memory block size, the memory block to be removed is automatically selected. However, when you specify a PA-start:size value to the mblock property, the specified memory blocks are explicitly removed.
ldm rm-mem [--auto-adj] size[unit] ldom ldm rm-mem mblock=PA-start:size[,PA-start:size[,...]] ldom
where:
--auto-adj specifies that the amount of memory to be removed from an active domain is automatically 256Mbyte-aligned, which might increase the requested memory size. If the domain is inactive, bound, or in a delayed reconfiguration, this option automatically aligns the resulting size of the domain by rounding up to the next 256-Mbyte boundary.
size is the size of memory to be removed from the logical domain.
unit is the unit of measurement. The default is bytes. If you want a different unit of measurement, specify one of the following (the unit is not case-sensitive):
G is gigabytes
K is kilobytes
M is megabytes
mblock=PA-start:size specifies one or more physical memory blocks to remove from a domain. PA-start specifies the starting physical address of the memory block in hexadecimal format. size is the size of the memory block, including a unit, to be removed from the domain.
ldom specifies the logical domain where the memory is to be removed.
This subcommand enables the domain to enter delayed reconfiguration mode. primary is the only valid value for ldom.
ldm start-reconf ldom
This subcommand cancels a delayed reconfiguration. primary is the only valid value for ldom.
ldm cancel-reconf ldom
This subcommand cancels a delayed reconfiguration (reconf), memory DR removal (memdr), or domain migration (migration) for a logical domain. For the reconf operation, primary is the only valid value for ldom.
ldm cancel-op migration ldom ldm cancel-op reconf ldom ldm cancel-op memdr ldom
This subcommand adds a PCI bus or device to a specified logical domain.
ldm add-io (bus | device | vf-name) ldom
where:
bus, device, and vf-name are a PCIe bus, a direct I/O-assignable device, and a PCIe SR-IOV virtual function, respectively. Although the operand can be specified as a device path or as a pseudonym, using the device pseudonym is recommended. The pseudonym is based on the ASCII label that is printed on the chassis to identify the corresponding I/O card slot and is platform specific.
The following are examples of the pseudonyms that are associated with the device names:
PCIe bus. The pci_0 pseudonym matches the pci@400 device path.
Direct I/O-assignable device. The PCIE1 pseudonym matches the pci@400/pci@0/pci@c device path.
PCIe SR-IOV virtual function. The /SYS/MB/NET0/IOVNET.PF0.VF0 pseudonym matches the pci@400/pci@2/pci@0/pci@6/network@0 device path.
The specified guest domain must be in the inactive or bound state. If you specify the primary domain, this command initiates a delayed reconfiguration.
ldom specifies the logical domain where the bus or device is to be added.
This subcommand modifies the current configuration of a virtual function by changing the property values or by passing new properties. This command can modify both the class-specific properties and the device-specific properties.
You can change most network class-specific properties without requiring a reboot of the primary domain. Changes to the mtu property do require a reboot. You must stop the domain if you change a virtual function MAC address.
All device-specific properties initiate a delayed reconfiguration so that those properties can be updated during the attach operation of the physical function device driver. As a result, the primary domain must be rebooted.
This command only succeeds when the physical function driver can successfully validate the resulting configuration.
ldm set-io [mac-addr=num] [alt-mac-addrs=[auto|num1,[auto|num2,...]]] [pvid=[port-vlan-id]] [vid=[vlan-id1,vlan-id2,...]] [mtu=size] [name=value...] vf-name
where:
name=value is the name-value pair of a property to set.
vf-name is the name of the virtual function.
This subcommand modifies the physical function configuration. Only the physical function device-specific properties are supported. Any change to the properties causes a delayed reconfiguration because the properties are applied during the attach operation of the physical function device driver.
The property values must be an integer or a string. Run the ldm list-io -d command to determine the property value type and whether a particular property can be set.
Note that the ldm set-io command succeeds only when the physical function driver successfully validates the resulting configuration.
ldm set-io name=value [name=value...] pf-name
where:
name=value is the name-value pair of a property to set.
pf-name is the name of the physical function.
This subcommand removes a PCI bus or device from a specified logical domain.
ldm rm-io (bus | device | vf-name) ldom
where:
bus, device, and vf-name are a PCIe bus, a direct I/O-assignable device, and a PCIe SR-IOV virtual function, respectively. Although the operand can be specified as a device path or as a pseudonym, using the device pseudonym is recommended. The pseudonym is based on the ASCII label that is printed on the chassis to identify the corresponding I/O card slot and is platform specific.
The following are examples of the pseudonyms that are associated with the device names:
PCIe bus. The pci_0 pseudonym matches the pci@400 device path.
Direct I/O-assignable device. The PCIE1 pseudonym matches the pci@400/pci@0/pci@c device path.
PCIe SR-IOV virtual function. The /SYS/MB/NET0/IOVNET.PF0.VF0 pseudonym matches the pci@400/pci@2/pci@0/pci@6/network@0 device path.
The specified guest domain must be in the inactive or bound state. If you specify the primary domain, this command initiates a delayed reconfiguration.
ldom specifies the logical domain where the bus or device is to be removed.
This subcommand adds a virtual switch to a specified logical domain.
ldm add-vsw [-q] [default-vlan-id=vlan-id] [pvid=port-vlan-id] [vid=vlan-id1,vlan-id2,...] [linkprop=phys-state] [mac-addr=num] [net-dev=device] [mode=sc] [mtu=size] [id=switch-id] [inter-vnet-link=on|off] vswitch-name ldom
where:
-q disables the validation of the path to the network device that is specified by the net-dev property. This option enables the command to run more quickly, especially if the logical domain is not fully configured.
default-vlan-id=vlan-id specifies the default VLAN to which a virtual switch and its associated virtual network devices belong to implicitly, in untagged mode. It serves as the default port VLAN ID (pvid) of the virtual switch and virtual network devices. Without this option, the default value of this property is 1. Normally, you would not need to use this option. It is provided only as a way to change the default value of 1.
pvid=port-vlan-id specifies the VLAN to which the virtual switch device needs to be a member, in untagged mode. This property also applies to the set-vsw subcommand. See Using VLAN Tagging in Oracle VM Server for SPARC 2.2 Administration Guide.
linkprop=phys-state specifies whether the virtual device reports its link status based on the underlying physical network device. When linkprop=phys-state is specified on the command line, the virtual device link status reflects the physical link state. By default, the virtual device link status does not reflect the physical link state.
vid=vlan-id specifies one or more VLANs to which a virtual network device or virtual switch needs to be a member, in tagged mode. This property also applies to the set-vsw subcommand. See Using VLAN Tagging in Oracle VM Server for SPARC 2.2 Administration Guide for more information.
mac-addr=num is the MAC address to be used by this switch. The number must be in standard octet notation, for example, 80:00:33:55:22:66. If you do not specify a MAC address, the switch is automatically assigned an address from the range of public MAC addresses allocated to the Logical Domains Manager.
net-dev=device is the path to the network device over which this switch operates. The system validates that the path references an actual network device unless the -q option is specified.
mode=sc enables virtual networking support for prioritized processing of Oracle Solaris Cluster heartbeat packets in a Logical Domains environment. Applications like Oracle Solaris Cluster need to ensure that high priority heartbeat packets are not dropped by congested virtual network and switch devices. This option prioritizes Oracle Solaris Cluster heartbeat frames and ensures that they are transferred in a reliable manner.
You must set this option when running Oracle Solaris Cluster in a Logical Domains environment and using guest domains as Oracle Solaris Cluster nodes. Do not set this option when you are not running Oracle Solaris Cluster software in guest domains because you could impact virtual network performance.
mtu=size specifies the maximum transmission unit (MTU) of a virtual switch device. Valid values are in the range of 1500-16000.
id=switch-id is the ID of a new virtual switch device. By default, ID values are generated automatically, so set this property if you need to match an existing device name in the OS.
inter-vnet-link=on|off specifies whether to assign a channel between each pair of virtual network devices that are connected to the same virtual switch. This behavior improves guest-to-guest performance. The default value is on.
vswitch-name is the unique name of the switch that is to be exported as a service. Clients (network) can attach to this service.
ldom specifies the logical domain in which to add a virtual switch.
This subcommand modifies the properties of a virtual switch that has already been added.
ldm set-vsw [-q] [pvid=port-vlan-id] [vid=vlan-id1,vlan-id2,...] [mac-addr=num] [net-dev=device] [linkprop=[phys-state]] [mode=[sc]] [mtu=size] [inter-vnet-link=[on|off]] vswitch-name
where:
-q disables the validation of the path to the network device that is specified by the net-dev property. This option enables the command to run more quickly, especially if the logical domain is not fully configured.
pvid=port-vlan-id specifies the VLAN to which the virtual switch device needs to be a member, in untagged mode. See Using VLAN Tagging in Oracle VM Server for SPARC 2.2 Administration Guide.
vid=vlan-id specifies one or more VLANs to which a virtual network device or virtual switch needs to be a member, in tagged mode. See Using VLAN Tagging in Oracle VM Server for SPARC 2.2 Administration Guide.
mac-addr=num is the MAC address used by the switch. The number must be in standard octet notation, for example, 80:00:33:55:22:66.
net-dev=device is the path to the network device over which this switch operates. The system validates that the path references an actual network device unless the -q option is specified.
linkprop=phys-state specifies whether the virtual device reports its link status based on the underlying physical network device. When linkprop=phys-state is specified on the command line, the virtual device link status reflects the physical link state. By default, the virtual device link status does not reflect the physical link state. The default situation occurs when the linkprop property is unspecified or when you run the ldm set-vsw command with the linkprop= argument.
mode=sc enables virtual networking support for prioritized processing of Oracle Solaris Cluster heartbeat packets in a Logical Domains environment. Applications like Oracle Solaris Cluster need to ensure that high priority heartbeat packets are not dropped by congested virtual network and switch devices. This option prioritizes Oracle Solaris Cluster heartbeat frames and ensures that they are transferred in a reliable manner.
mode= (left blank) stops special processing of heartbeat packets.
You must set this option when running Oracle Solaris Cluster in a Logical Domains environment and using guest domains as Oracle Solaris Cluster nodes. Do not set this option when you are not running Oracle Solaris Cluster software in guest domains because you could impact virtual network performance.
mtu=size specifies the maximum transmission unit (MTU) of a virtual switch device. Valid values are in the range of 1500-16000.
inter-vnet-link=on|off specifies whether to assign a channel between each pair of virtual network devices that are connected to the same virtual switch. This behavior improves guest-to-guest performance. The default value is on.
vswitch-name is the unique name of the switch that is to exported as a service. Clients (network) can be attached to this service.
This subcommand removes a virtual switch.
ldm rm-vsw [-f] vswitch-name
where:
-f attempts to force the removal of a virtual switch. The removal might fail.
vswitch-name is the name of the switch that is to be removed as a service.
This subcommand adds a virtual network device to the specified logical domain.
ldm add-vnet [mac-addr=num] [mode=hybrid] [pvid=port-vlan-id] [vid=vlan-id1,vlan-id2,...] [linkprop=phys-state] [id=network-id] [mtu=size] if-name vswitch-name ldom
where:
mac-addr=num is the MAC address for this network device. The number must be in standard octet notation, for example, 80:00:33:55:22:66.
mode=hybrid requests the system to use NIU Hybrid I/O on this vnet if possible. If it is not possible, the system reverts to virtual I/O. This hybrid mode is considered a delayed reconfiguration if set on an active vnet on a control domain. See Using NIU Hybrid I/O in Oracle VM Server for SPARC 2.2 Administration Guide.
pvid=port-vlan-id specifies the VLAN to which the virtual network device needs to be a member, in untagged mode. See Using VLAN Tagging in Oracle VM Server for SPARC 2.2 Administration Guide.
vid=vlan-id specifies one or more VLANs to which a virtual network device needs to be a member, in tagged mode. See Using VLAN Tagging in Oracle VM Server for SPARC 2.2 Administration Guide.
mtu=size specifies the maximum transmission unit (MTU) of a virtual network device. Valid values are in the range of 1500-16000.
linkprop=phys-state specifies whether the virtual network device reports its link status based on the underlying physical network device. When linkprop=phys-state is specified on the command line, the virtual network device link status reflects the physical link state. By default, the virtual network device link status does not reflect the physical link state.
id=network-id is the ID of a new virtual network device. By default, ID values are generated automatically, so set this property if you need to match an existing device name in the OS.
if-name is a unique interface name to the logical domain, which is assigned to this virtual network device instance for reference on subsequent set-vnet or rm-vnet subcommands.
vswitch-name is the name of an existing network service (virtual switch) to which to connect.
ldom specifies the logical domain to which to add the virtual network device.
This subcommand sets options for a virtual network device in the specified logical domain.
ldm set-vnet [mac-addr=num] [vswitch=vswitch-name] [mode=[hybrid]] [pvid=port-vlan-id] [linkprop=[phys-state]] [vid=vlan-id1,vlan-id2,...] [mtu=size] if-name ldom
where:
mac-addr=num is the MAC address for this network device. The number must be in standard octet notation, for example, 80:00:33:55:22:66.
vswitch=vswitch-name is the name of an existing network service (virtual switch) to which to connect.
mode=hybrid enables NIU Hybrid I/O operations on this vnet. This option is considered a delayed reconfiguration if set on an active vnet on a control domain. Leave the mode= argument blank to disable NIU Hybrid I/O.
pvid=port-vlan-id specifies the VLAN to which the virtual network device needs to be a member, in untagged mode. See Using VLAN Tagging in Oracle VM Server for SPARC 2.2 Administration Guide.
linkprop=phys-state specifies whether the virtual device reports its link status based on the underlying physical network device. When linkprop=phys-state is specified on the command line, the virtual device link status reflects the physical link state. By default, the virtual device link status does not reflect the physical link state. The default situation occurs when the linkprop property is unspecified or when you run the ldm set-vnet command with the linkprop= argument.
vid=vlan-id specifies one or more VLANs to which a virtual network device needs to be a member, in tagged mode. See Using VLAN Tagging in Oracle VM Server for SPARC 2.2 Administration Guide.
mtu=size specifies the maximum transmission unit (MTU) of a virtual network device. Valid values are in the range of 1500-16000.
if-name is the unique interface name assigned to the virtual network device that you want to set.
ldom specifies the logical domain in which to modify the virtual network device.
This subcommand removes a virtual network device from the specified logical domain.
ldm rm-vnet [-f] if-name ldom
where:
-f attempts to force the removal of a virtual network device from a logical domain. The removal might fail.
if-name is the unique interface name assigned to the virtual network device that you want to remove.
ldom specifies the logical domain from which to remove the virtual network device.
This subcommand adds a virtual disk server to the specified logical domain.
ldm add-vds service-name ldom
where:
service-name is the service name for this instance of the virtual disk server. The service-name must be unique among all virtual disk server instances on the server.
ldom specifies the logical domain in which to add the virtual disk server.
This subcommand removes a virtual disk server.
ldm rm-vds [-f] service-name
where:
-f attempts to force the removal of a virtual disk server. The removal might fail.
service-name is the unique service name for this instance of the virtual disk server.
 | Caution - The -f option attempts to unbind all clients before removal, which might cause loss of disk data if writes are in progress. |
This subcommand adds a device to a virtual disk server. The device can be an entire disk, a slice on a disk, a file, or a disk volume. See Chapter 7, Using Virtual Disks, in Oracle VM Server for SPARC 2.2 Administration Guide.
ldm add-vdsdev [-f] [-q] [options={ro,slice,excl}] [mpgroup=mpgroup] backend volume-name@service-name
where:
-f attempts to force the creation of an additional virtual disk server when specifying a block device path that is already part of another virtual disk server. If specified, the -f option must be the first in the argument list.
-q disables the validation of the virtual disk back end that is specified by the backend operand. This option enables the command to run more quickly, especially if the logical domain or the back end is not fully configured.
options= are as follows:
ro – Specifies read-only access
slice – Exports a back end as a single slice disk
excl – Specifies exclusive disk access
Omit the options= argument to have the default values of disk, not exclusive, and read/write. If you add the options= argument, you must specify one or more of the options for a specific virtual disk server device. Separate two or more options with commas and no spaces, such as ro,slice,excl.
mpgroup=mpgroup is the disk multipath group name used for virtual disk failover support. You can assign the virtual disk several redundant paths in case the link to the virtual disk server device currently in use fails. To do this, you would group multiple virtual disk server devices (vdsdev) into one multipath group (mpgroup), all having the same mpgroup name. When a virtual disk is bound to any virtual disk server device in a multipath group, the virtual disk is bound to all the virtual disk server devices that belong to the mpgroup.
backend is the location where data of a virtual disk are stored. The back end can be a disk, a disk slice, a file, a volume (including ZFS, Solaris Volume Manager, or VxVM), or any disk pseudo device. The disk label can be SMI VTOC, EFI, or no label at all. A back end appears in a guest domain either as a full disk or as single slice disk, depending on whether the slice option is set when the back end is exported from the service domain. When adding a device, the volume-name must be paired with the backend. The system validates that the location specified by backend exists and can be used as a virtual disk back end unless the -q option is specified.
volume-name is a unique name that you must specify for the device being added to the virtual disk server. The volume-name must be unique for this virtual disk server instance because this name is exported by this virtual disk server to the clients for adding. When adding a device, the volume-name must be paired with the backend.
service-name is the name of the virtual disk server to which to add this device.
This subcommand sets options for a virtual disk server. See the Oracle VM Server for SPARC 2.2 Administration Guide.
ldm set-vdsdev [-f] options=[{ro,slice,excl}] [mpgroup=mpgroup] volume-name@service-name
where:
-f removes the read-only restriction when multiple volumes in the same logical domain are sharing an identical block device path in read-only mode (option=ro). If specified, the -f option must be the first in the argument list.
options= are as follows:
ro – Specifies read-only access
slice – Exports a back end as a single slice disk
excl – Specifies exclusive disk access
Leave the options= argument blank to turn off any previous options specified. You can specify all or a subset of the options for a specific virtual disk server device. Separate two or more options with commas and no spaces, such as ro,slice,excl.
mpgroup=mpgroup is the disk multipath group name used for virtual disk failover support. You can assign the virtual disk several redundant paths in case the link to the virtual disk server device currently in use fails. To do this, you would group multiple virtual disk server devices (vdsdev) into one multipath group (mpgroup), all having the same mpgroup name. When a virtual disk is bound to any virtual disk server device in a multipath group, the virtual disk is bound to all the virtual disk server devices that belong to the mpgroup.
volume-name is the name of an existing volume exported by the service named by service-name.
service-name is the name of the virtual disk server being modified.
This subcommand removes a device from a virtual disk server.
ldm rm-vdsdev [-f] volume-name@service-name
where:
-f attempts to force the removal of the virtual disk server device. The removal might fail.
volume-name is the unique name for the device being removed from the virtual disk server.
service-name is the name of the virtual disk server from which to remove this device.
| Caution - Without the -f option, the rm-vdsdev subcommand does not allow a virtual disk server device to be removed if the device is busy. Using the -f option can cause data loss for open files. |
This subcommand adds a virtual disk to the specified logical domain. An optional timeout property allows you to specify a timeout for a virtual disk if it cannot establish a connection with the virtual disk server.
ldm add-vdisk [timeout=seconds] [id=disk-id] disk-name volume-name@service-name ldom
where:
timeout=seconds is the number of seconds for establishing a connection between a virtual disk client (vdc) and a virtual disk server (vds). If there are multiple virtual disk (vdisk) paths, then the vdc can try to connect to a different vds, and the timeout ensures that a connection to any vds is established within the specified amount of time.
Omit the timeout= argument or set timeout=0 to have the virtual disk wait indefinitely.
id=disk-id is the ID of a new virtual disk device. By default, ID values are generated automatically, so set this property if you need to match an existing device name in the OS.
disk-name is the name of the virtual disk.
volume-name is the name of the existing virtual disk server device to which to connect.
service-name is the name of the existing virtual disk server to which to connect.
ldom specifies the logical domain in which to add the virtual disk.
This subcommand sets options for a virtual disk in the specified logical domain. An optional timeout property allows you to specify a timeout for a virtual disk if it cannot establish a connection with the virtual disk server.
ldm set-vdisk [timeout=seconds] [volume=volume-name@service-name] disk-name ldom
where:
timeout=seconds is the number of seconds for establishing a connection between a virtual disk client (vdc) and a virtual disk server (vds). If there are multiple virtual disk (vdisk) paths, then the vdc can try to connect to a different vds, and the timeout ensures that a connection to any vds is established within the specified amount of time.
Set timeout=0 to disable the timeout.
Do not specify a timeout= argument to have the virtual disk wait indefinitely.
volume=volume-name is the name of the virtual disk server device to which to connect. service-name is the name of the virtual disk server to which to connect.
disk-name is the name of the existing virtual disk.
ldom specifies the existing logical domain where the virtual disk was previously added.
This subcommand removes a virtual disk from the specified logical domain.
ldm rm-vdisk [-f] disk-name ldom
where:
-f attempts to force the removal of the virtual disk. The removal might fail.
disk-name is the name of the virtual disk to be removed.
ldom specifies the logical domain from which to remove the virtual disk.
This subcommand adds a virtual data plane channel service to the specified logical domain. This subcommand should only be used in a Netra Data Plane Software (NDPS) environment.
ldm add-vdpcs vdpcs-service-name ldom
where:
vdpcs-service-name is the name of the virtual data plane channel service that is to be added.
ldom specifies the logical domain to which to add the virtual data plane channel service.
This subcommand removes a virtual data plane channel service. This subcommand should only be used in a Netra Data Plane Software (NDPS) environment.
ldm rm-vdpcs [-f] vdpcs-service-name
where:
-f attempts to force the removal of the virtual data plane channel service. The removal might fail.
vdpcs-service-name is the name of the virtual data plane channel service that is to be removed.
This subcommand adds a virtual data plane channel client to the specified logical domain. This subcommand should only be used in a Netra Data Plane Software (NDPS) environment.
ldm add-vdpcc vdpcc-name vdpcs-service-name ldom
where:
vdpcc-name is the unique name of the virtual data plane channel service client.
vdpcs-service-name is the name of the virtual data plane channel service to which to connect this client.
ldom specifies the logical domain to which to add the virtual data plane channel client.
This subcommand removes a virtual data plane channel client from the specified logical domain. This subcommand should only be used in a Netra Data Plane Software (NDPS) environment.
ldm rm-vdpcc [-f] vdpcc-name ldom
where:
-f attempts to force the removal of the virtual data plane channel client. The removal might fail.
vdpcc-name is the unique name assigned to the virtual data plane channel client that is to be removed.
ldom specifies the logical domain from which to remove the virtual data plane channel client.
This subcommand adds a virtual console concentrator to the specified logical domain.
ldm add-vcc port-range=x-y vcc-name ldom
where:
port-range=x-y is the range of TCP ports to be used by the virtual console concentrator for console connections.
vcc-name is the name of the virtual console concentrator that is to be added.
ldom specifies the logical domain to which to add the virtual console concentrator.
This subcommand sets options for a specific virtual console concentrator.
ldm set-vcc port-range=x-y vcc-name
where:
port-range=x-y is the range of TCP ports to be used by the virtual console concentrator for console connections. Any modified port range must encompass all the ports assigned to clients of the concentrator.
vcc-name is the name of the virtual console concentrator that is to be set.
This subcommand removes a virtual console concentrator from the specified logical domain.
ldm rm-vcc [-f] vcc-name
where:
-f attempts to force the removal of the virtual console concentrator. The removal might fail.
vcc-name is the name of the virtual console concentrator that is to be removed.
| Caution - The -f option attempts to unbind all clients before removal, which might cause loss of data if writes are in progress. |
This subcommand sets a specific port number and group in the specified logical domain. You can also set the attached console's service. This subcommand can be used only when a domain is inactive.
ldm set-vcons [port=[port-num]] [group=group] [service=vcc-server] ldom
where:
port=port-num is the specific port to use for this console. Leave the port-num blank to have the Logical Domains Manager automatically assign the port number.
group=group is the new group to which to attach this console. The group argument allows multiple consoles to be multiplexed onto the same TCP connection. Refer to the Oracle Solaris OS vntsd(1M) man page for more information about this concept. When a group is specified, a service must also be specified.
service=vcc-server is the name for the existing virtual console concentrator that should handle the console connection. A service must be specified when a group is specified.
ldom specifies the logical domain in which to set the virtual console concentrator.
The PCIe single-root I/O virtualization (SR-IOV) standard enables the efficient sharing of PCIe devices among I/O domains. This standard is implemented in the hardware to achieve near-native I/O performance. SR-IOV creates a number of virtual functions that are virtualized instances of the physical device or function. The virtual functions are directly assigned to I/O domains so that they can share the associated physical device and perform I/O without CPU and hypervisor overhead.
PCIe physical functions have complete access to the hardware and provide the SR-IOV capability to create, configure, and manage virtual functions. A PCIe component on the system board or a PCIe plug-in card can provide one or more physical functions. An Oracle Solaris driver interacts with the physical functions that provide access to the SR-IOV features.
PCIe virtual functions contain the resources that are necessary for data movement. An I/O domain that has a virtual function can access hardware and perform I/O directly by means of an Oracle Solaris virtual function driver. This behavior avoids the overhead and latency that is involved in the virtual I/O feature by removing any bottlenecks in the communication path between the applications that run in the I/O domain and the physical I/O device in the root domain.
Some of these commands require that you specify an identifier for a physical function or virtual function as follows:
pf-name ::= pf-pseudonym | pf-path vf-name ::= vf-pseudonym | vf-path
Use the pseudonym form when referring to a corresponding device. This is the form of the name that is shown in the NAME column of the ldm list-io output. When you run the ldm list-io -l command, the path form of the name appears in the output. The ldm list-io -p output shows the pseudonym form as the value of the alias= token and the path form as the value of the dev= token.
Create a Virtual FunctionThis subcommand creates a virtual function from a specified physical function by incrementing the number of virtual functions in the specified physical function by one. The new virtual function is assigned the highest number in the sequence of virtual function numbers. When you create a new virtual function, a delayed reconfiguration is initiated for the primary domain.
Network class virtual functions must have a MAC address assigned, which is assigned by default. To override the default MAC address value, specify another value for the mac-addr property.
You can also set class-specific properties and device-specific properties when you create a virtual function. This command succeeds only when the physical function driver successfully validates the resulting configuration. By default, a new virtual function is not assigned to any domain. The virtual function can only be assigned (bound) to an I/O domain after the primary domain is rebooted and the virtual function is instantiated in the hardware. Plan ahead by determining whether you want to create multiple virtual functions. If you do, create them one after the other to avoid performing multiple reboots.
The device-specific properties depend on the properties that are exported by the physical function driver. For more information, use the ldm list-io -d command. When the command is successful, you see a message about a delayed reconfiguration.
ldm create-vf [mac-addr=num] [alt-mac-addrs=auto|num1,[auto|num2,...]] [pvid=port-vlan-id] [vid=vlan-id1,vlan-id2,...] [mtu=size] [name=value...] pf-name
where:
mac-addr=num is the primary MAC address of the virtual function
alt-mac-addrs=auto|num1,[auto|num2,...] is a comma-separated list of alternate MAC addresses. Valid values are numeric MAC addresses and the auto keyword, which can be used one or more times to request that the system generate an alternate MAC address. The auto keyword can be mixed with numeric MAC addresses.
pvid=port-vlan-id is the port VLAN ID (no default value)
vid=vlan-id1,vlan-id2... is a comma-separated list of integer VLAN IDs.
mtu=size is the maximum transmission unit (in bytes) for the virtual function.
name=value is the name-value pair of a property to specify.
pf-name is the name of the physical function.
This subcommand destroys a virtual function from the specified physical function. This command succeeds only if the following are true:
The specified virtual function is not currently assigned to any domain.
The specified virtual function is the last virtual function in the corresponding physical function.
The resulting configuration is successfully validated by the physical function driver.
A successful operation triggers a delayed reconfiguration, as the change to the number of virtual functions can only be done as part of rebooting. See the create-vf subcommand for more information.
ldm destroy-vf vf-name
where vf-name is the name of the virtual function.
This subcommand adds one or more variables for a logical domain.
ldm add-var var-name=[value]... ldom
where:
var-name=value is the name-value pair of a variable to add. The value is optional.
ldom specifies the logical domain in which to add the variable.
This subcommand sets variables for a logical domain.
ldm set-var var-name=[value]... ldom
where:
var-name=value is the name-value pair of a variable to set. The value is optional.
ldom specifies the logical domain in which to set the variable.
Note - Leaving value blank, sets var-name to no value.
This subcommand removes a variable for a logical domain.
ldm rm-var var-name... ldom
where:
var-name is the name of a variable to remove.
ldom specifies the logical domain from which to remove the variable.
This subcommand starts one or more logical domains.
ldm start -a ldm start -i file ldm start ldom...
where:
-a starts all bound logical domains.
-i file specifies an XML configuration file to use in starting the logical domain.
ldom specifies one or more logical domains to start.
This subcommand stops one or more running logical domains. The subcommand sends a uadmin request to the logical domain if the Oracle Solaris OS is booted. to stop a domain in a more “graceful” manner, perform a shutdown or init operation in the domain that you want to stop. See the shutdown(1M) and init(1M) man page.
ldm stop [-f] -a ldm stop [-f] ldom...
where:
-f attempts to force a running logical domain to stop. Use only if the domain cannot be stopped by any other means.
-a stops all running logical domains except the control domain.
ldom specifies one or more running logical domains to stop.
This subcommand panics the Oracle Solaris OS on a specified logical domain, which provides a back trace and crash dump if you configure the Oracle Solaris OS to do that. The dumpadm(1M) command provides the means to configure the crash dump.
ldm panic ldom
ldom specifies the logical domain to panic.
Provide Help InformationThis subcommand provides usage for all subcommands or the subcommand that you specify. You can also use the ldm command alone to provide usage for all subcommands.
ldm --help [subcommand]
subcommand specifies the ldm subcommand about which you want usage information.
Provide Version InformationThis subcommand provides version information.
ldm --version ldm -VBind Resources to a Domain
This subcommand binds, or attaches, configured resources to a logical domain.
ldm bind-dom [-f] [-q] -i file ldm bind-dom [-f] [-q] ldom
where:
-f attempts to force the binding of the domain even if invalid network or disk back-end devices are detected.
-q disables the validation of network or disk back-end devices so that the command runs more quickly.
-i file specifies an XML configuration file to use in binding the logical domain.
ldom specifies the logical domain to which to bind resources.
This subcommand releases resources bound to configured logical domains.
ldm unbind-dom ldom
ldom specifies the logical domain from which to unbind resources.
This subcommand adds a logical domain configuration, either based on the currently active configuration or on a previously autosaved configuration. The configuration is stored on the service processor (SP).
ldm add-config config-name ldm add-config -r autosave-name [new-config-name]
where:
config-name is the name of the logical domain configuration to add.
-r autosave-name applies the autosave configuration data to one of the following:
Configuration on the SP that has the same name
Newly created configuration, new-config-name, which does not exist on the SP
If the target configuration does not exist on the SP, a configuration of that name is created and saved to the SP based on the contents of the corresponding autosave configuration. After the autosave configuration data is applied, those autosave files are deleted from the control domain. If autosave-name does not represent the currently selected configuration, or if new-config-name is specified, the state of the current configuration on the SP and any autosave files for it on the control domain are unaffected.
To recover an autosave configuration that is known to be corrupted, you must specify -r new-config-name. You are not permitted to overwrite an existing configuration with one that is known to be corrupted.
new-config-name is the name of the logical domain configuration to add.
This subcommand enables you to specify a logical domain configuration to use. The configuration is stored on the SP.
ldm set-config config-name
config-name is the name of the logical domain configuration to use.
The default configuration name is factory-default. To specify the default configuration, use the following:
ldm set-config factory-defaultRemove a Logical Domain Configuration
This subcommand removes a logical domain configuration that is stored on the SP, as well as any corresponding autosave configuration from the control domain.
ldm rm-config [-r] config-name
where:
-r only removes autosave configurations from the control domain.
config-name is the name of the logical domain configuration to remove.
This subcommand lists logical domains and their states. If you do not specify a logical domain, all logical domains are listed.
ldm ls-dom [-e] [-l] [-o format] [-p][ [ldom...]
where:
-e generates an extended listing containing services and devices that are automatically set up, that is, not under your control.
-l generates a long listing.
-o limits the output format to one or more of the following subsets. If you specify more than one format, delimit each format by a comma with no spaces.
console – Output contains the virtual console (vcons) and virtual console concentrator (vcc) service.
core – Output contains information about cores, core ID and physical CPU set.
cpu – Output contains information about the CPU thread (vcpu), physical CPU (pcpu), and core ID (cid).
crypto – Cryptographic unit output contains the Modular Arithmetic Unit (mau) and any other supported cryptographic unit, such as the Control Word Queue (CWQ).
disk – Output contains the virtual disk (vdisk) and virtual disk server (vds).
domain – Output contains the variables (var), host ID (hostid), domain state, flags, universally unique identifier (UUID), software state, utilization percentage, a slave's master domains, and the master domain's failure policy.
memory – Output contains memory.
network – Output contains the media access control (mac) address , virtual network switch (vsw), and virtual network (vnet) device.
physio – Physical input/output contains the peripheral component interconnect (pci) and network interface unit (niu).
resmgmt – Output contains DRM policy information, indicates which policy is currently running, and indicates whether the whole-core, max-core, and threading constraints are enabled.
serial – Output contains the virtual logical domain channel (vldc) service, virtual logical domain channel client (vldcc), virtual data plane channel client (vdpcc), and virtual data plane channel service (vdpcs).
status – Output contains the status of a migrating domain.
-p generates the list in a parseable, machine-readable format.
ldom is the name of the logical domain for which to list state information.
This subcommand lists bindings for logical domains. If no logical domains are specified, all logical domains are listed.
ldm ls-bindings [-e] [-p] [ldom...]
where:
-e generates an extended listing containing services and devices that are automatically set up, that is, not under your control.
-p generates the list in a parseable, machine-readable format.
ldom is the name of the logical domain for which you want binding information.
This subcommand lists all the services exported by logical domains. If no logical domains are specified, all logical domains are listed.
ldm ls-services [-e] [-p] [ldom...]
where:
-e generates an extended listing containing services and devices that are automatically set up, that is, not under your control.
-p generates the list in a parseable, machine-readable format.
ldom is the name of the logical domain for which you want services information.
This subcommand lists the constraints for the creation of one or more logical domains. If no logical domains are specified, all logical domains are listed.
ldm ls-constraints [-x] [ldom...] ldm ls-constraints [-e] [-p] [ldom...]
where:
-x writes the constraint output in XML format to the standard output (stdout) format. This output can be used as a backup.
ldom is the name of the logical domain for which you want to list constraints.
-e generates an extended listing containing services and devices that are automatically set up, that is, not under your control.
-p writes the constraint output in a parseable, machine-readable form.
This subcommand lists either free (unbound) resources or all server resources. The default is to list all free resources.
ldm ls-devices [-a] [-p] [core] [cpu] [crypto] [memory] [io]
where:
-a lists all server resources, bound and unbound.
-p writes the constraint output in a parseable, machine-readable form.
core lists information about cores, the core ID and physical CPU set, and specifies which CPUs in the core are still unallocated.
cpu lists CPU thread and physical CPU resources.
crypto lists only the modular arithmetic unit resources.
memory lists only memory resources.
io lists only input/output resources, such as a PCI bus, a network, or direct I/O-assignable devices.
In the power management column (PM) or field (pm=), yes means that the CPU thread is power-managed, and no means that the CPU thread is powered on. It is assumed that 100 percent-free CPUs are power-managed by default.
List I/O DevicesThis subcommand lists the I/O devices that are configured on the system. The list of devices includes I/O buses (including NIUs) and direct I/O-assignable devices.
The output is divided into the following sections:
I/O bus information. The IO column lists the device path of the bus or network device, and the PSEUDONYM column shows the associated pseudonym for the bus or network device. The DOMAIN column indicates the domain to which the device is currently bound.
Direct I/O-assignable devices. The PCIE column lists the device path of the device, and the PSEUDONYM column shows the associated pseudonym for the device.
The STATUS column applies to slots that accept plug-in cards. The value is either OCC (occupied), EMP (empty), or UNK (unknown). Slots that represent on-board devices always have the status of OCC. If the root domain does not support the direct I/O (DIO) feature, the slot status is UNK.
The DOMAIN column indicates the domain to which the device is currently bound. If this field is empty, the corresponding device is not bound. If this field shows a hyphen character (-), the device cannot currently be removed from the primary domain for one of the following reasons:
The OS that is running on the root domain does not support the DIO feature. In this case, the STATUS field for all the devices shown in the PCIE column of the output is UNK.
The corresponding slot is empty so that no PCIe option card is present. Note that empty slots cannot be assigned to direct I/O domains.
The root complex to which the device is connected is assigned to another domain by means of a split PCI configuration. In this case, the STATUS column for the device is UNK.
ldm list-io [-l] [-p] [pf-name] ldm list-io -d pf-name
where:
-l lists information about subdevices that are hosted by direct I/O-assignable devices. Note that this output indicates which devices will be loaned with the direct I/O-assignable device to the receiving domain. The subdevice names cannot be used for command input.
-p writes the output in a parseable, machine-readable form.
-d pf-name lists information about the specified physical function.
pf-name is the name of the physical function.
This subcommand lists the logical domain configurations stored on the service processor (SP).
ldm ls-config [-r [autosave-name]]
-r [autosave-name] lists those configurations for which autosave files exist on the control domain. If autosave-name is specified, it only reports on autosave-name. The output also notes whether an autosave file is newer than the corresponding SP configuration.
Note - When a delayed reconfiguration is pending, the configuration changes are immediately autosaved. As a result, if you run the ldm ls-config -r command, the autosave configuration is shown as being newer than the current configuration.
This subcommand lists one or more variables for a logical domain. To list all variables for a domain, leave the var-name blank.
ldm ls-var [var-name...] ldom
where:
var-name is the name of the variable to list. If you do not specify any name, all variables will be listed for the domain.
ldom is the name of the logical domain for which to list one or more variables.
This subcommand enables you to add a resource management policy for one or more logical domains. A resource management policy consists of optional properties and their values.
You can enable a resource management policy in an active domain that supports CPU DR as long as power management is not using the elastic policy.
ldm add-policy [enable=yes|no] [priority=value] [attack=value] [decay=value] [elastic-margin=value] [sample-rate=value] [tod-begin=hh:mm[:ss]] [tod-end=hh:mm[:ss]] [util-lower=percent] [util-upper=percent] [vcpu-min=value] [vcpu-max=value] name=policy-name ldom...
where:
The properties are described in the Properties section.
ldom specifies the logical domain for which to add a resource management policy.
This subcommand enables you to modify a resource management policy for one or more logical domains by specifying values for optional properties.
ldm set-policy [enable=[yes|no]] [priority=[value]] [attack=[value]] [decay=[value]] [elastic-margin=[value]] [sample-rate=[value]] [tod-begin=[hh:mm:ss]] [tod-end=[hh:mm:ss]] [util-lower=[percent]] [util-upper=[percent]] [vcpu-min=[value]] [vcpu-max=[value]] name=policy-name ldom...
where:
The properties are described in the Properties section.
ldom specifies the logical domain for which to modify the resource management policy.
This subcommand enables you to remove a resource management policy from a logical domain by specifying one or more policy names.
ldm remove-policy [name=]policy-name... ldom
where:
The name property specifies the name of the resource management policy, policy-name.
ldom specifies the logical domain on which to remove the resource management policy.
This subcommand enables you to use an existing configuration to configure one or more guest domains, the control domain, or both types of domains. The ldm init-system command takes an XML file (such as the output of ldm ls-constraints -x) as input, configures the specified domains, and reboots the control domain. Run this command with the factory default configuration.
ldm init-system [-frs] -i file
where:
-i file specifies the XML configuration file to use to create the logical domain.
-f skips the factory-default configuration check and continues irrespective of what was already configured on the system.
Use the -f option with caution. ldm init-system assumes that the system is in the factory-default configuration, and so directly applies the changes that are specified by the XML file. Using -f when the system is in a configuration other than the factory default will likely result in a system that is not configured as specified by the XML file. One or more changes might fail to be applied to the system depending on the combination of changes in the XML file and the initial configuration.
-r reboots the system after configuration.
-s restores only the virtual services configuration (vds, vcc, and vsw).
Example 1 Create Default Services
Set up the three default services, virtual disk server, virtual switch, and virtual console concentrator so that you can export those services to the guest domains.
# ldm add-vds primary-vds0 primary # ldm add-vsw net-dev=nxge0 primary-vsw0 primary # ldm add-vcc port-range=5000-5100 primary-vcc0 primary
Example 2 List Services
You can list services to ensure they have been created correctly or to see what services you have available.
# ldm ls-services primary VCC NAME LDOM PORT-RANGE primary-vcc0 primary 5000-5100 VSW NAME LDOM MAC NET-DEV DEVICE DEFAULT-VLAN-ID PVID VID MODE primary-vsw0 primary 00:14:4f:f9:68:d0 nxge0 switch@0 1 1 VDS NAME LDOM VOLUME OPTIONS MPGROUP DEVICE primary-vds0 primary
Example 3 Set Up the Control Domain Initially
The control domain, named primary, is the initial domain that is present when you install the Logical Domains Manager. The control domain has a full complement of resources, and those resources depend on what server you have. Set only those resources you want the control domain to keep so that you can allocate the remaining resources to the guest domains. Then save the configuration on the service processor. You must reboot so the changes take effect.
If you want to enable networking between the control domain and the other domains, you must plumb the virtual switch on the control domain. You must enable the virtual network terminal server daemon, vntsd(1M), to use consoles on the guest domains.
# ldm start-reconf primary # ldm set-crypto 1 primary # ldm set-vcpu 4 primary # ldm set-mem 4G primary # ldm add-config initial # shutdown -y -g0 -i6 # ifconfig -a # ifconfig vsw0 plumb # ifconfig nxge0 down unplumb # ifconfig vsw0 IP-of-nxge0 netmask netmask-of-nxge0 broadcast + up # svcadm enable vntsd
Example 4 List Bindings
You can list bindings to see if the control domain has the resources you specified, or what resources are bound to any domain.
# ldm ls-bindings primary NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME primary active -t-cv 4 4G 12% 11m MAC 08:00:90:11:11:10 VCPU VID PID UTIL STRAND 0 0 18% 100% 1 1 13% 100% 2 2 9.8% 100% 3 3 5.4% 100% MEMORY RA PA SIZE 0x4000000 0x4000000 4G IO DEVICE PSEUDONYM OPTIONS pci@400 pci_0 pci@500 pci_1 pci@400/pci@0/pci@9 PCIE2 pci@400/pci@0/pci@9 MB/SASHBA pci@500/pci@0/pci@8 MB/NET0 VCC NAME PORT-RANGE primary-vcc0 5000-5100 VSW NAME MAC NET-DEV DEVICE MODE primary-vsw0 00:14:4f:f9:68:d0 nxge0 switch@0 prog,promisc VDS NAME VOLUME OPTIONS DEVICE primary-vds0
Example 5 Create a Logical Domain
Ensure that you have the resources to create the desired guest domain configuration, add the guest domain, add the resources and devices that you want the domain to have, set boot parameters to tell the system how to behave on startup, bind the resources to the domain, and save the guest domain configuration in an XML file for backup. You also might want to save the primary and guest domain configurations on the SC. Then you can start the domain, find the TCP port of the domain, and connect to it through the default virtual console service.
# ldm ls-devices # ldm add-dom ldg1 # ldm add-vcpu 4 ldg1 # ldm add-mem 1g ldg1 # ldm add-vnet vnet1 primary-vsw0 ldg1 # ldm add-vdsdev /dev/dsk/c0t1d0s2 vol1@primary-vds0 # ldm add-vdisk vdisk1 vol1@primary-vds0 ldg1 # ldm set-var auto-boot\?=false ldg1 # ldm set-var boot-device=vdisk1 ldg1 # ldm bind-dom ldg1 # ldm ls-constraints -x ldg1 > ldg1.xml # ldm add-config ldg1_4cpu_1G # ldm start ldg1 # ldm ls -l ldg1 # telnet localhost 5000
Example 6 Use One Terminal for Many Guest Domains
Normally, each guest domain you create has its own TCP port and console. Once you have created the first guest domain (ldg1 in this example), you can use the ldm set-vcons command to attach all the other domains (second domain is ldg2 in this example) to the same console port. Note that the set-vcons subcommand works only on an inactive domain.
# ldm set-vcons group=ldg1 service=primary-vcc0 ldg2
If you use the ldm ls -l command after performing the set-vcons commands on all guest domains except the first, you can see that all domains are connected to the same port. See the vntsd(1M) man page for more information about using consoles.
Example 7 Add a Virtual PCI Bus to a Logical Domain
I/O domains are a type of service domain that have direct ownership of and direct access to physical I/O devices. The I/O domain then provides the service to the guest domain in the form of a virtual I/O device. This example shows how to add a virtual PCI bus to a logical domain.
# ldm add-io pci@7c0 ldg1
Example 8 Add Virtual Data Plane Channel Functionality for Netra Only
If your server has a Netra Data Plane Software (NDPS) environment, you might want to add virtual data plane channel functionality. First, you would add a virtual data plane channel service (primary-vdpcs0, for example) to the service domain, in this case, the primary domain.
# ldm add-vdpcs primary-vdpcs0 primary
Now that you have added the service to the service domain (primary), you can add the virtual data plane channel client (vdpcc1) to a guest domain (ldg1).
# add-vdpcc vdpcc1 primary-vdpcs0 ldg1
Example 9 Cancel Delayed Reconfiguration Operations for a Control Domain
A delayed reconfiguration operation blocks configuration operations on all other domains. There might be times when you want to cancel delayed configuration operations for a control domain. For example, you might do this so that you can perform other configuration commands on that domain or other domains. With this command, you can undo the delayed reconfiguration operation and do other configuration operations on this or other domains.
# ldm cancel-op reconf primary
Example 10 Migrate a Domain
You can migrate a logical domain to another machine. This example shows a successful migration.
# ldm migrate ldg1 root@dt90-187:ldg Target password:
Example 11 List Configurations
The following examples show how to view the configurations. The first command shows the configurations that are stored on the SP. The second command shows the configurations on the SP as well as information about the autosave configurations on the control domain.
# ldm ls-config factory-default 3guests [current] data1 reconfig_primary split1 # ldm ls-config -r 3guests [newer] data1 [newer] reconfig_primary split1 unit
Both the current 3guests configuration and the data1 configuration have autosaved changes that have not been saved to the SP. If the system is powercycled while in this state, the Logical Domains Manager would perform the 3guests autosave recovery based on the specified policy. The autosave recovery action is taken for 3guests because it is marked as current.
The reconfig_primary and split1 autosave configurations are identical to the versions on the SP, not newer versions.
The unit configuration only exists as an autosave configuration on the control domain. There is no corresponding configuration for unit on the SP. This situation might occur if the configuration was lost from the SP. A configuration can be lost if the SP is replaced or if a problem occurred with the persistent version of the configuration on the SP. Note that using the rm-config command to explicitly remove a configuration also removes the autosave version on the control domain. As a result, no remnants of the configuration remain on either the control domain or on the SP.
Example 12 List I/O Devices
The following example lists the I/O devices on the system. The first section shows information about the PCIe bus on the primary domain. The second section shows information about the direct I/O-assignable devices.
# ldm ls-io IO PSEUDONYM DOMAIN -- --------- ------ pci@400 pci_0 primary pci@500 pci_1 primary PCIE PSEUDONYM STATUS DOMAIN ---- --------- ------ ------ pci@400/pci@0/pci@c PCIE1 EMP - pci@400/pci@0/pci@9 PCIE2 OCC ldg1 pci@400/pci@0/pci@d PCIE3 OCC ldg2 pci@400/pci@0/pci@8 MB/SASHBA OCC primary pci@500/pci@0/pci@9 PCIE0 EMP - pci@500/pci@0/pci@d PCIE4 OCC ldg2 pci@500/pci@0/pci@c PCIE5 OCC ldg1 pci@500/pci@0/pci@8 MB/NET0 OCC primary
The following exit values are returned:
Successful completion.
An error occurred.
See the attributes(5) man page for a description of the following attributes.
|
dumpadm(1M), ifconfig(1M), shutdown(1M), vntsd(1M), attributes(5)
|