Oracle® VM Server User's Guide Release 2.2 Part Number E15444-04 |
|
|
PDF · Mobi · ePub |
This Appendix contains references for the Oracle VM Server and Oracle VM Agent command-line tools. The command-line interfaces in this Appendix are:
The ovs-agent command-line tool enables you to configure, and control Oracle VM Agent. Enter parameters to the ovs-agent command-line tool in the format
service ovs-agent {option}
If you shut down or restart the Oracle VM Agent on an HA-enabled Oracle VM Server on which guests are running you are prompted to:
Migrate or Power Off the guest(s) using Oracle VM Manager. When the guests have been migrated or Powered Off, the Oracle VM Agent shuts down.
Shut down the guest(s) and shut down Oracle VM Agent.
Cancel the shutdown operation.
See Chapter 3, "Oracle VM Agent" for examples on using the ovs-agent command-line tool.
Starts Oracle VM Agent.
# service ovs-agent start
Stops Oracle VM Agent.
# service ovs-agent stop
Stops and restarts Oracle VM Agent.
# service ovs-agent restart
Displays information on the status of the Oracle VM Agent daemon.
# service ovs-agent status
Starts the Oracle VM Agent interactive configuration script.
# service ovs-agent configure
The virt-install command-line tool creates paravirtualized guests and hardware virtualized guests. virt-install can be used as an interactive shell, or all parameters can be given at the same time. Enter multiple parameters to the virt-install command-line tool in the format:
virt-install [option ...]
This section contains a brief explanation of some of the more common virt-install options. For full documentation, use the virt-install -h
command.
Chapter 4, "Creating a Guest Virtual Machine" discusses using the virt-install tool.
Displays the virt-install command parameters and their purpose.
# virt-install -h
Sets the name of the guest instance.
# virt-install -nMyGuest
Sets the memory to allocate for a guest instance in Megabytes.
# virt-install --ram=256
Sets the UUID (Universally Unique Identifier) for the guest. If none is given, a random UUID is generated.
# virt-install -u
Sets the number of virtual CPUs to configure for the guest.
# virt-install --vcpus=2
Sets the file to use as the disk image.
# virt-install --file=/home/myhome/myimage
Sets the size of the disk image (if it does not exist) in Gigabytes.
# virt-install -s2
Do not use sparse files for disks. This option may be significantly slower when creating guests.
Sets the fixed MAC address for the guest; if none or RANDOM is given, a random address is used.
# virt-install --mac=RANDOM
Sets the bridge to connect guest NIC to. If none is given, attempts to determine the default.
Use VNC (Virtual Network Computing) for graphics support.
# virt-install --vnc
Sets the port to use for VNC connections.
# virt-install --vncport=5900
Use SDL (Simple DirectMedia Layer) for graphics support.
# virt-install --sdl
Do not use a graphical console for the guest.
# virt-install --nographics
Do not automatically connect to the guest console.
# virt-install --noautoconsole
Set up keyboard mapping for the graphical console. If none is given, the keymap is automatically set to the local keymap.
# virt-install --de
Use kernel acceleration capabilities.
# virt-install --accelerate
Connect to hypervisor with URI.
# virt-install --connect=test:///default
Sets the guest as being a fully virtualized guest.
# virt-install -v
Sets the file to use a virtual CD-ROM device for fully a virtualized guest.
Sets the operating system type for a fully virtualized guest. Possible values are windows
, unix
, other
, and linux
.
# virt-install --os-type=windows
Sets the operating system variant for a fully virtualized guest, for example, rhel5
, win2k
, or vista
. This parameter should be used with the os-type
parameter.
The following table lists the possible values available for os-variant
for each os-type
option.
Possible values for os-type=windows | Possible values for os-type=unix | Possible values for os-type=other | Possible values for os-type=linux |
---|---|---|---|
win2k3 | solaris9 | netware6 | generic24 |
win2k | solaris10 | generic | generic26 |
vista | freebsd6 | netware4 | rhel2.1 |
winxp | openbsd4 | msdos | fedora7_64 |
netware5 | el5_64 | ||
fedora6 | |||
fedora7 | |||
fedora5 | |||
centos5_64 | |||
generic26_64 | |||
centos5 | |||
sles10 | |||
sles10_64 | |||
el4_64 | |||
rhel4 | |||
rhel5 | |||
rhel4_64 | |||
rhel3 | |||
fedora6_64 | |||
rhel5_64 | |||
fedora5_64 | |||
el4 | |||
el5 |
Note:
Not all operating system variants are supported by Oracle for use with Oracle products, but are made available for your convenience.# virt-install --os-type=windows --os-variant=winxp
Disables APIC (Advanced Programmable Interrupt Controller) for a fully virtualized guest. Overrides the value set in --os-type
and --os-variant
.
Sets the CPU architecture to simulate.
# virt-install --arch=x86
Sets the guest as being a paravirtualized guest.
Sets the installation source for a paravirtualized guest, for example, nfs:host:/path, http://host/path, or ftp://host/path.
# virt-install -lhttp://example.com/path
Sets the virtual network interface type for hardware virtualized guests. The netfront driver is a paravirtualized driver which can be used with a paravirtualized guest, or with a hardware virtualized guest with the proper paravirtualized drivers installed. The ioemu driver is a hardware virtualized driver, and can only be used with a hardware virtualized guest. Both drivers contain the device emulation code to support hardware virtualized guests.
For hardware virtualized guests, type
can be either ioemu
or netfront
. The default is ioemu
.
You cannot use this parameter for paravirtualized guests. For paravirtualized guests, the default is netfront
and cannot be changed.
# virt-install --vif-type=ioemu
Any additional arguments to pass to the installer with a paravirtualized guest.
Prints debugging information.
The Oracle VM Server management command-line management tool xm, creates, destroys, manages and migrates guests.
This section contains a brief explanation of some of the more common xm commands. For full documentation, use the xm help --long
command.
The xm command-line tool requires the xend daemon to be started.
Enter parameters to the xm command-line tool in the format:
xm [option] [argument]
See Chapter 5, "Domain Monitoring and Administration" for examples on using the xm command-line tool.
Creates a new virtual block device. This triggers a hotplug event for the guest.
The domain-id parameter is the domain identifier.
The be-dev parameter is the device in the back-end domain (usually dom0) to be exported. Can be specified as a physical partition (phy:sda7), or as a file mounted as loopback (file://path/to/loop.iso).
The fe-dev parameter specifies how the device is presented to the guest domain. Can be specified as either a symbolic name, such as /dev/hdc for common devices, or by a device identifier, such as 0x1400 (/dev/hdc device identifier in hexadecimal).
The mode parameter sets the access mode for the device from the guest domain. Supported modes are w
(read/write) or r
(read-only).
The bedomain-id parameter is the back-end domain hosting the device. Defaults to dom0.
For example, to mount an ISO as a disk:
# xm block-attach 02 file://path/to/dsl-2.0RC2.iso /dev/hdc ro
This mounts the ISO as /dev/hdc in the guest domain as a read only device. This may not be detected as a CD-ROM by the guest, but mounting /dev/hdc gives you access to the ISO.
Detaches a virtual block devices from a domain.
The device-id parameters may be the symbolic name or the numeric device identifier given to the device by dom0. Run the xm block-list
command to determine the device identifier.Detaching the device requires the co-operation of the domain. If the domain fails to release the device (perhaps because the domain is hung or is still using the device), the detach fails. The --force parameter forces the device to detach, but may cause IO errors in the domain.
# xm block-detach 02 51776
Lists the virtual block devices for a domain.
# xm block-list 02 --long
Attaches to a domain's console.
# xm console 02
Creates a domain based on the entries in the config-file
.
Entering the -c
parameter attaches to the domain's console when the domain is created and started. You can also enter name value pairs to override variables in the config-file
using the name=value
parameter.
# xm -c /home/myhome/myconfig
Immediately terminates a domain.
# xm destroy 02
Displays message buffer logs similar in format to the equivalent to the dmesg command in the Linux kernel.
The --clear
parameter clears the message buffer.
Displays help on the xm command, and its options.
The --long
option displays full help on xm commands, grouped by function.
Enter a command name as an option to the xm command to get help only on that command.
# xm help --long create
Displays information about the host computer.
# xm info
Displays information on all the running domains.
The --long
option displays full information on running domains.
Enter the domain-id
as an option to the xm command to get information on only that domain, or a set of domains.
# xm list --long 02
Displays logs similar in format to the equivalent for the Linux kernel. The log file is located at /var/log/xend.log.
# xm log
Allocates the maximum memory available to the domain in MBs. Overrides the maxmen
setting in the configuration file.
# xm mem-max 02 2048
Allocates domain memory up to the value of the maxmen
setting in the guest configuration file or the mem-max {domain-id} {MBs} setting. Overrides the memory
setting in the guest configuration file.
# xm mem-set 02 1024
Migrates a domain to another computer.
The domain-id
parameter is the domain to migrate.
The host
parameter is the target computer.
The --live
parameter migrates the domain without shutting down the domain.
The --resource
parameter sets the maximum amount of Megabytes to be used.
# xm migrate 02 example.com --live
Adds a domain to Oracle VM Server domain management.
You can set domain creation parameters with a number of command-line options, a Python script (with the --defconfig
parameter), or an SXP configuration file (the --config
parameter).
You can set configuration variables with name=value
pairs, for example vmid=3
sets vmid to 3.
The config-file
parameter is the location of the domain configuration file.
The option
parameter is one or more of the following:
[-h | --help]
Displays help on the command.
[--help-config]
Prints the available configuration variables for the configuration script.
[-q | --quiet]
Quiet.
[--path=path]
Searches the location given in path
for configuration scripts. The value of path
is a colon-separated directory list.
[-f=file | --defconfig=file]
Uses the given Python configuration script. The script is loaded after arguments have been processed. Each command-line option sets a configuration variable named after its long option name, and these variables are placed in the environment of the script before it is loaded. Variables for options that may be repeated have list values. Other variables can be set using name=value
on the command-line. After the script is loaded, values that were not set on the command-line are replaced by the values set in the script.
[-F=file | --config=file]
Sets the domain configuration to use SXP. SXP is the underlying configuration format used by Xen. SXP configurations can be hand-written or generated from Python configuration scripts, using the --dryrun
option to print the configuration.
[-n | --dryrun]
Prints the resulting configuration in SXP, but does not create the domain.
[-x | --xmldryrun]
Prints the resulting configuration in XML, but does not create the domain.
[-s | --skipdtd]
Skips DTD checking and XML checks before domain creation. This option is experimental and may slow down the creation of domains.
[-p | --paused]
Leaves the domain paused after it is created.
[-c | --console_autoconnect]
Connects to the console after the domain is created.
# xm new /home/myhome/myconfig
Pauses the execution of a domain.
# xm pause 02
Reboots a domain.
The --all
parameter reboots all domains.
The --wait
parameter waits for the domain to reboot before returning control to the console.
# xm reboot --wait 02
Restores a domain from a saved state.
# xm restore /home/myhome/statefile
Saves a domain state so it can be restored at a later date.
# xm save 02 /home/myhome/statefile
Shuts down a domain gracefully.
The -a
parameter shuts down all domains.
The -w
parameter waits for the domain to shut down before returning control to the console.
# xm shutdown -w 02
Displays real time monitoring information of the host and domains.
# xm top
Unpauses a paused domain.
# xm unpause 02
Sets the number of virtual CPUs for the domain, up to the value of the vcpus
setting in the guest configuration file.
# xm vcpu-set 02 8
You can set which virtual CPUs are used with the vcpu_avail parameter in the configuration file. See Section C.3, "Virtual CPU Configuration File Parameters".
The Physical to Virtual (P2V) conversion utility enables you to convert a computer's operating system (Linux and Windows) and applications to an Oracle VM hardware virtualized guest image. The P2V utility is included on the Oracle VM Server CD. You can access the P2V utility by restarting a computer with the Oracle VM Server CD. The Oracle VM Server startup screen is displayed. At the boot:
prompt, enter:
linux p2v
You can use a P2V kickstart file to automate creation of hardware virtualized guest images from physical computers. This section discusses the options and parameters of the P2V kickstart file.
The P2V utility converts disks on the computer to virtual disk images. The virtual disk images are created as IDE disks (hda, hdb, hdc, hdd, and so on) on the guest, using the original disk names. When you use a P2V kickstart file, up to four disks are automatically deployed in the guest. Any extra disks are converted and added to the guest configuration file (vm.cfg), although they are not deployed. To deploy the additional disks in the guest, edit the guest configuration file, remove the comments from the disk entries, and map the additional disks to SCSI device names, for example, sda, sdb, and sdc. The boot disk must always be mapped to device hda. Any files on the guest which contain references to these devices must also be changed, for example, the /etc/fstab file may contain references to /dev/hda1, /dev/sda1, and so on.
When you use a P2V kickstart file, at least one network interface must use DHCP. This is required for the computer running the P2V utility to read the kickstart file over the network. The network configuration for this network interface cannot be modified from the kickstart file.
If you want the P2V utility's web server to listen using a network interface other than the one used to initiate the kickstart session, the network configuration (DHCP or static IP address) for that network interface can be specified in the kickstart file.
A number of screens may be displayed prior to the P2V utility starting with a kickstart file. You can suppress these screens to fully automate the P2V utility. Prior to the P2V utility starting, you may see up to four screens:
The following sections give examples on how to suppress these screens.
Suppressing the P2V Network Configuration Screen
To suppress the P2V Network Configuration screen, supply the ethernet device on the command line, for example:
linux p2v ks=http://example.com/ks.cfg ksdevice=eth0
Suppressing the Language Selection Screen
To suppress the Language selection screen, supply the language kickstart parameter, for example:
lang en_US.UTF-8
Suppressing the Keyboard Selection Screen
To suppress the Keyboard selection screen, supply the keyboard kickstart parameter, for example:
keyboard us
Suppressing the Installation Source Screen
To suppress the Installation source screen, supply the source kickstart parameter, for example:
cdrom
An example P2V kickstart file follows:
p2v cdrom lang en_US.UTF-8 keyboard us target --ovmmanager network --device eth0 --bootproto dhcp diskimage --device /dev/sda --type IDE vm_options --name myGuest --mem 1024 --vcpus 1 --consolepasswd mypassword
For more examples and information on using P2V kickstart files, see "Converting a Linux or Windows Host" in Chapter 10, "Converting Hosts and Non-Oracle VM Virtual Machines".
The following parameters are accepted in a P2V kickstart file.
Indicates the kickstart file is intended to automate a P2V conversion. This parameter is required in order to perform an automated P2V conversion and should be supplied at the Oracle VM Server boot:
prompt instead of install
, update
, or rescue
. It accepts no parameters.
Sets the end destination for the guest image.
The option
parameter can only contain the following:
--ovmmanager
Sets the P2V utility to operate in HTTPS server mode to transfer the guest image to a running instance of Oracle VM Manager.
Denotes a disk to be included in the guest image. The P2V utility uses device mapper-based snapshotting to copy the disk as a system-*.img file on the target computer. There may be multiple diskimage
directives in a P2V kickstart file, each resulting in a disk image in the guest image. The --device
parameter must always be used with the diskimage
directive to indicate which device should be imaged.
The option
parameter is one or more of the following:
--device path
The device to image. path
must be the full path to the device. For example:
diskimage --device /dev/sda
--type [IDE | SCSI | LVM | MDRAID]
Sets the type of disk. Must be one of IDE
, SCSI
, LVM
, or MDRAID
. Devices /dev/hda, /dev/hdb, /dev/hdc, and /dev/hdd should be IDE
. Devices /dev/sd[a-zz] should be SCSI
. A logical volume should be LVM
. Devices /dev/md[a-zz] should be MDRAID
. For example:
diskimage --device /dev/hda --type IDE
Configures network information for the computer.
The option
parameter is one or more of the following:
--bootproto [dhcp | bootp | static]
Sets the method by which the network configuration is determined. Must be dhcp
, bootp
, or static
. The default is dhcp
. bootp
and dhcp
are treated as the same.
dhcp
uses a DHCP server to obtain the networking configuration, for example:
network --bootproto dhcp
static
requires all the necessary networking information. As the name implies, this information is static and is used during and after the installation. The entry for static networking is more complex, as you must include all network configuration information on one line. You must specify the IP address, netmask, gateway, and nameserver, for example:
network --bootproto static --ip 10.0.2.15 --netmask 255.255.255.0 --gateway 10.0.2.254 --nameserver 10.0.2.1
The static
method has the following restrictions:
All static networking configuration information must be specified on one line; you cannot wrap lines using a backslash.
You can only specify one nameserver.
--ip ipaddress
The IP address for the computer.
--gateway ipaddress
The IP address for the default gateway.
--nameserver ipaddress
The IP address for the primary nameserver.
--netmask netmask
The netmask for the computer.
Sets the configuration options for the guest.
--name name
The name of the guest.
--mem size
The memory allocation for the guest in Mb.
--vcpus number
The number of VCPUs for the guest.
--consolepasswd password
The console password for the guest. For example:
vm_options --name myGuest --mem 1024 --vcpus 1 --consolepasswd mypassword