1 gnt-instance(8) Ganeti | Version @GANETI_VERSION@
2 =================================================
7 gnt-instance - Ganeti instance administration
12 **gnt-instance** {command} [arguments...]
17 The **gnt-instance** command is used for instance administration in
23 Creation/removal/querying
24 ~~~~~~~~~~~~~~~~~~~~~~~~~
30 | {-t|\--disk-template {diskless \| file \| plain \| drbd \| rbd}}
31 | {\--disk=*N*: {size=*VAL*[,spindles=*VAL*] \| adopt=*LV*}[,options...]
32 | \| {size=*VAL*,provider=*PROVIDER*}[,param=*value*... ][,options...]
33 | \| {-s|\--os-size} *SIZE*}
34 | [\--no-ip-check] [\--no-name-check] [\--no-conflicts-check]
35 | [\--no-start] [\--no-install]
36 | [\--net=*N* [:options...] \| \--no-nics]
37 | [{-B|\--backend-parameters} *BEPARAMS*]
38 | [{-H|\--hypervisor-parameters} *HYPERVISOR* [: option=*value*... ]]
39 | [{-O|\--os-parameters} *param*=*value*... ]
40 | [\--file-storage-dir *dir\_path*] [\--file-driver {loop \| blktap}]
41 | {{-n|\--node} *node[:secondary-node]* \| {-I|\--iallocator} *name*}
42 | {{-o|\--os-type} *os-type*}
43 | [\--submit] [\--print-job-id]
47 Creates a new instance on the specified host. The *instance* argument
48 must be in DNS, but depending on the bridge/routing setup, need not be
49 in the same network as the nodes in the cluster.
51 The ``disk`` option specifies the parameters for the disks of the
52 instance. The numbering of disks starts at zero, and at least one disk
53 needs to be passed. For each disk, either the size or the adoption
54 source needs to be given. The size is interpreted (when no unit is
55 given) in mebibytes. You can also use one of the suffixes *m*, *g* or
56 *t* to specify the exact the units used; these suffixes map to
57 mebibytes, gibibytes and tebibytes. Each disk can also take these
58 parameters (all optional):
61 How many spindles (physical disks on the node) the disk should span.
64 The access mode. Either ``ro`` (read-only) or the default ``rw``
68 This option specifies a name for the disk, which can be used as a disk
69 identifier. An instance can not have two disks with the same name.
72 The LVM volume group. This works only for LVM and DRBD devices.
75 This options specifies a different VG for the metadata device. This
76 works only for DRBD devices
78 When creating ExtStorage disks, also arbitrary parameters can be passed,
79 to the ExtStorage provider. Those parameters are passed as additional
80 comma separated options. Therefore, an ExtStorage disk provided by
81 provider ``pvdr1`` with parameters ``param1``, ``param2`` would be
82 passed as ``--disk 0:size=10G,provider=pvdr1,param1=val1,param2=val2``.
84 When using the ``adopt`` key in the disk definition, Ganeti will
85 reuse those volumes (instead of creating new ones) as the
86 instance's disks. Ganeti will rename these volumes to the standard
87 format, and (without installing the OS) will use them as-is for the
88 instance. This allows migrating instances from non-managed mode
89 (e.g. plain KVM with LVM) to being managed via Ganeti. Please note that
90 this works only for the \`plain' disk template (see below for
93 Alternatively, a single-disk instance can be created via the ``-s``
94 option which takes a single argument, the size of the disk. This is
95 similar to the Ganeti 1.2 version (but will only create one disk).
97 The minimum disk specification is therefore ``--disk 0:size=20G`` (or
98 ``-s 20G`` when using the ``-s`` option), and a three-disk instance
99 can be specified as ``--disk 0:size=20G --disk 1:size=4G --disk
102 The minimum information needed to specify an ExtStorage disk are the
103 ``size`` and the ``provider``. For example:
104 ``--disk 0:size=20G,provider=pvdr1``.
106 The ``--no-ip-check`` skips the checks that are done to see if the
107 instance's IP is not already alive (i.e. reachable from the master
110 The ``--no-name-check`` skips the check for the instance name via
111 the resolver (e.g. in DNS or /etc/hosts, depending on your setup).
112 Since the name check is used to compute the IP address, if you pass
113 this option you must also pass the ``--no-ip-check`` option.
115 If you don't want the instance to automatically start after
116 creation, this is possible via the ``--no-start`` option. This will
117 leave the instance down until a subsequent **gnt-instance start**
120 The NICs of the instances can be specified via the ``--net``
121 option. By default, one NIC is created for the instance, with a
122 random MAC, and set up according the the cluster level NIC
123 parameters. Each NIC can take these parameters (all optional):
126 either a value or 'generate' to generate a new unique MAC
129 specifies the IP address assigned to the instance from the Ganeti
130 side (this is not necessarily what the instance will use, but what
131 the node expects the instance to use). Note that if an IP in the
132 range of a network configured with **gnt-network**\(8) is used,
133 and the NIC is not already connected to it, this network has to be
134 passed in the **network** parameter if this NIC is meant to be
135 connected to the said network. ``--no-conflicts-check`` can be used
136 to override this check. The special value **pool** causes Ganeti to
137 select an IP from the the network the NIC is or will be connected to.
140 specifies the connection mode for this NIC: routed, bridged or
144 in bridged or openvswitch mode specifies the interface to attach
145 this NIC to, in routed mode it's intended to differentiate between
146 different routing tables/instance groups (but the meaning is
147 dependent on the network script, see **gnt-cluster**\(8) for more
148 details). Note that openvswitch support is also hypervisor
152 derives the mode and the link from the settings of the network
153 which is identified by its name. If the network option is chosen,
154 link and mode must not be specified. Note that the mode and link
155 depend on the network-to-nodegroup connection, thus allowing
156 different nodegroups to be connected to the same network in
160 this option specifies a name for the NIC, which can be used as a NIC
161 identifier. An instance can not have two NICs with the same name.
164 Of these "mode" and "link" are NIC parameters, and inherit their
165 default at cluster level. Alternatively, if no network is desired for
166 the instance, you can prevent the default of one NIC with the
167 ``--no-nics`` option.
169 The ``-o (--os-type)`` option specifies the operating system to be
170 installed. The available operating systems can be listed with
171 **gnt-os list**. Passing ``--no-install`` will however skip the OS
172 installation, allowing a manual import if so desired. Note that the
173 no-installation mode will automatically disable the start-up of the
174 instance (without an OS, it most likely won't be able to start-up
177 The ``-B (--backend-parameters)`` option specifies the backend
178 parameters for the instance. If no such parameters are specified, the
179 values are inherited from the cluster. Possible parameters are:
182 the maximum memory size of the instance; as usual, suffixes can be
183 used to denote the unit, otherwise the value is taken in mebibytes
186 the minimum memory size of the instance; as usual, suffixes can be
187 used to denote the unit, otherwise the value is taken in mebibytes
190 the number of VCPUs to assign to the instance (if this value makes
191 sense for the hypervisor)
194 whether the instance is considered in the N+1 cluster checks
195 (enough redundancy in the cluster to survive a node failure)
198 ``True`` or ``False``, whether the instance must be failed over
199 (shut down and rebooted) always or it may be migrated (briefly
202 Note that before 2.6 Ganeti had a ``memory`` parameter, which was the
203 only value of memory an instance could have. With the
204 ``maxmem``/``minmem`` change Ganeti guarantees that at least the minimum
205 memory is always available for an instance, but allows more memory to be
206 used (up to the maximum memory) should it be free.
208 The ``-H (--hypervisor-parameters)`` option specified the hypervisor
209 to use for the instance (must be one of the enabled hypervisors on the
210 cluster) and optionally custom parameters for this instance. If not
211 other options are used (i.e. the invocation is just -H *NAME*) the
212 instance will inherit the cluster options. The defaults below show the
213 cluster defaults at cluster creation time.
215 The possible hypervisor options are as follows:
218 Valid for the Xen HVM and KVM hypervisors.
220 A string value denoting the boot order. This has different meaning
221 for the Xen HVM hypervisor and for the KVM one.
223 For Xen HVM, The boot order is a string of letters listing the boot
224 devices, with valid device letters being:
238 The default is not to set an HVM boot order, which is interpreted
241 For KVM the boot order is either "floppy", "cdrom", "disk" or
242 "network". Please note that older versions of KVM couldn't netboot
243 from virtio interfaces. This has been fixed in more recent versions
244 and is confirmed to work at least with qemu-kvm 0.11.1. Also note
245 that if you have set the ``kernel_path`` option, that will be used
246 for booting, and this setting will be silently ignored.
249 Valid for the Xen HVM and PVM hypervisors.
251 Relevant to non-pvops guest kernels, in which the disk device names
252 are given by the host. Allows one to specify 'xvd', which helps run
253 Red Hat based installers, driven by anaconda.
256 Valid for the KVM hypervisor.
258 The path to a floppy disk image to attach to the instance. This
259 is useful to install Windows operating systems on Virt/IO disks
260 because you can specify here the floppy for the drivers at
264 Valid for the Xen HVM and KVM hypervisors.
266 The path to a CDROM image to attach to the instance.
269 Valid for the KVM hypervisor.
271 The path to a second CDROM image to attach to the instance.
272 **NOTE**: This image can't be used to boot the system. To do that
273 you have to use the 'cdrom\_image\_path' option.
276 Valid for the Xen HVM and KVM hypervisors.
278 This parameter determines the way the network cards are presented
279 to the instance. The possible options are:
281 - rtl8139 (default for Xen HVM) (HVM & KVM)
282 - ne2k\_isa (HVM & KVM)
283 - ne2k\_pci (HVM & KVM)
289 - paravirtual (default for KVM) (HVM & KVM)
292 Valid for the Xen HVM hypervisor.
294 This parameter specifies the vif type of the nic configuration
295 of the instance. Unsetting the value leads to no type being specified
296 in the configuration. Note that this parameter only takes effect when
297 the 'nic_type' is not set. The possible options are:
303 Valid for the Xen HVM and KVM hypervisors.
305 This parameter determines the way the disks are presented to the
306 instance. The possible options are:
308 - ioemu [default] (HVM & KVM)
317 Valid for the KVM hypervisor.
319 This parameter determines the way the cdroms disks are presented
320 to the instance. The default behavior is to get the same value of
321 the earlier parameter (disk_type). The possible options are:
332 Valid for the Xen HVM and KVM hypervisors.
334 Specifies the address that the VNC listener for this instance
335 should bind to. Valid values are IPv4 addresses. Use the address
336 0.0.0.0 to bind to all available interfaces (this is the default)
337 or specify the address of one of the interfaces on the node to
338 restrict listening to that interface.
341 Valid for the KVM hypervisor.
343 A boolean option that controls whether the VNC connection is
347 Valid for the KVM hypervisor.
349 If ``vnc_tls`` is enabled, this options specifies the path to the
350 x509 certificate to use.
353 Valid for the KVM hypervisor.
356 Valid for the KVM hypervisor.
358 Specifies the address or interface on which the SPICE server will
359 listen. Valid values are:
361 - IPv4 addresses, including 0.0.0.0 and 127.0.0.1
362 - IPv6 addresses, including :: and ::1
363 - names of network interfaces
365 If a network interface is specified, the SPICE server will be bound
366 to one of the addresses of that interface.
369 Valid for the KVM hypervisor.
371 Specifies which version of the IP protocol should be used by the
374 It is mainly intended to be used for specifying what kind of IP
375 addresses should be used if a network interface with both IPv4 and
376 IPv6 addresses is specified via the ``spice_bind`` parameter. In
377 this case, if the ``spice_ip_version`` parameter is not used, the
378 default IP version of the cluster will be used.
380 spice\_password\_file
381 Valid for the KVM hypervisor.
383 Specifies a file containing the password that must be used when
384 connecting via the SPICE protocol. If the option is not specified,
385 passwordless connections are allowed.
387 spice\_image\_compression
388 Valid for the KVM hypervisor.
390 Configures the SPICE lossless image compression. Valid values are:
399 spice\_jpeg\_wan\_compression
400 Valid for the KVM hypervisor.
402 Configures how SPICE should use the jpeg algorithm for lossy image
403 compression on slow links. Valid values are:
409 spice\_zlib\_glz\_wan\_compression
410 Valid for the KVM hypervisor.
412 Configures how SPICE should use the zlib-glz algorithm for lossy image
413 compression on slow links. Valid values are:
419 spice\_streaming\_video
420 Valid for the KVM hypervisor.
422 Configures how SPICE should detect video streams. Valid values are:
428 spice\_playback\_compression
429 Valid for the KVM hypervisor.
431 Configures whether SPICE should compress audio streams or not.
434 Valid for the KVM hypervisor.
436 Specifies that the SPICE server must use TLS to encrypt all the
437 traffic with the client.
440 Valid for the KVM hypervisor.
442 Specifies a list of comma-separated ciphers that SPICE should use
443 for TLS connections. For the format, see man **cipher**\(1).
446 Valid for the KVM hypervisor.
448 Enables or disables passing mouse events via SPICE vdagent.
451 Valid for the KVM hypervisor.
453 This parameter determines the emulated cpu for the instance. If this
454 parameter is empty (which is the default configuration), it will not
457 Be aware of setting this parameter to ``"host"`` if you have nodes
458 with different CPUs from each other. Live migration may stop working
461 For more information please refer to the KVM manual.
464 Valid for the Xen HVM and KVM hypervisors.
466 A boolean option that specifies if the hypervisor should enable
467 ACPI support for this instance. By default, ACPI is disabled.
470 Valid for the Xen HVM and KVM hypervisors.
472 A boolean option that specifies if the hypervisor should enable
473 PAE support for this instance. The default is false, disabling PAE
477 Valid for the Xen HVM hypervisor.
479 A boolean option that specifies if the hypervisor should enable
480 viridian (Hyper-V) for this instance. The default is false,
481 disabling viridian support.
484 Valid for the Xen HVM and KVM hypervisors.
486 A boolean option that specifies if the instance should be started
487 with its clock set to the localtime of the machine (when true) or
488 to the UTC (When false). The default is false, which is useful for
489 Linux/Unix machines; for Windows OSes, it is recommended to enable
493 Valid for the Xen PVM and KVM hypervisors.
495 This option specifies the path (on the node) to the kernel to boot
496 the instance with. Xen PVM instances always require this, while for
497 KVM if this option is empty, it will cause the machine to load the
498 kernel from its disks (and the boot will be done accordingly to
502 Valid for the Xen PVM and KVM hypervisors.
504 This options specifies extra arguments to the kernel that will be
505 loaded. device. This is always used for Xen PVM, while for KVM it
506 is only used if the ``kernel_path`` option is also specified.
508 The default setting for this value is simply ``"ro"``, which
509 mounts the root disk (initially) in read-only one. For example,
510 setting this to single will cause the instance to start in
514 Valid for the Xen PVM and KVM hypervisors.
516 This option specifies the path (on the node) to the initrd to boot
517 the instance with. Xen PVM instances can use this always, while
518 for KVM if this option is only used if the ``kernel_path`` option
519 is also specified. You can pass here either an absolute filename
520 (the path to the initrd) if you want to use an initrd, or use the
521 format no\_initrd\_path for no initrd.
524 Valid for the Xen PVM and KVM hypervisors.
526 This options specifies the name of the root device. This is always
527 needed for Xen PVM, while for KVM it is only used if the
528 ``kernel_path`` option is also specified.
530 Please note, that if this setting is an empty string and the
531 hypervisor is Xen it will not be written to the Xen configuration
535 Valid for the KVM hypervisor.
537 This boolean option specifies whether to emulate a serial console
538 for the instance. Note that some versions of KVM have a bug that
539 will make an instance hang when configured to use the serial console
540 unless a connection is made to it within about 2 seconds of the
541 instance's startup. For such case it's recommended to disable this
542 option, which is enabled by default.
545 Valid for the KVM hypervisor.
547 This integer option specifies the speed of the serial console.
548 Common values are 9600, 19200, 38400, 57600 and 115200: choose the
549 one which works on your system. (The default is 38400 for historical
550 reasons, but newer versions of kvm/qemu work with 115200)
553 Valid for the KVM hypervisor.
555 The disk cache mode. It can be either default to not pass any
556 cache option to KVM, or one of the KVM cache modes: none (for
557 direct I/O), writethrough (to use the host cache but report
558 completion to the guest only when the host has committed the
559 changes to disk) or writeback (to use the host cache and report
560 completion as soon as the data is in the host cache). Note that
561 there are special considerations for the cache mode depending on
562 version of KVM used and disk type (always raw file under Ganeti),
563 please refer to the KVM documentation for more details.
566 Valid for the KVM hypervisor.
568 The security model for kvm. Currently one of *none*, *user* or
569 *pool*. Under *none*, the default, nothing is done and instances
570 are run as the Ganeti daemon user (normally root).
572 Under *user* kvm will drop privileges and become the user
573 specified by the security\_domain parameter.
575 Under *pool* a global cluster pool of users will be used, making
576 sure no two instances share the same user on the same node. (this
577 mode is not implemented yet)
580 Valid for the KVM hypervisor.
582 Under security model *user* the username to run the instance
583 under. It must be a valid username existing on the host.
585 Cannot be set under security model *none* or *pool*.
588 Valid for the KVM hypervisor.
590 If *enabled* the -enable-kvm flag is passed to kvm. If *disabled*
591 -disable-kvm is passed. If unset no flag is passed, and the
592 default running mode for your kvm binary will be used.
595 Valid for the KVM hypervisor.
597 This option passes the -mem-path argument to kvm with the path (on
598 the node) to the mount point of the hugetlbfs file system, along
599 with the -mem-prealloc argument too.
602 Valid for the KVM hypervisor.
604 This boolean option determines whether to run the KVM instance in a
607 If it is set to ``true``, an empty directory is created before
608 starting the instance and its path is passed via the -chroot flag
609 to kvm. The directory is removed when the instance is stopped.
611 It is set to ``false`` by default.
614 Valid for the KVM hypervisor.
616 The maximum amount of time (in ms) a KVM instance is allowed to be
617 frozen during a live migration, in order to copy dirty memory
618 pages. Default value is 30ms, but you may need to increase this
619 value for busy instances.
621 This option is only effective with kvm versions >= 87 and qemu-kvm
625 Valid for the Xen, KVM and LXC hypervisors.
627 The processes belonging to the given instance are only scheduled
628 on the specified CPUs.
630 The format of the mask can be given in three forms. First, the word
631 "all", which signifies the common case where all VCPUs can live on
632 any CPU, based on the hypervisor's decisions.
634 Second, a comma-separated list of CPU IDs or CPU ID ranges. The
635 ranges are defined by a lower and higher boundary, separated by a
636 dash, and the boundaries are inclusive. In this form, all VCPUs of
637 the instance will be mapped on the selected list of CPUs. Example:
638 ``0-2,5``, mapping all VCPUs (no matter how many) onto physical CPUs
641 The last form is used for explicit control of VCPU-CPU pinnings. In
642 this form, the list of VCPU mappings is given as a colon (:)
643 separated list, whose elements are the possible values for the
644 second or first form above. In this form, the number of elements in
645 the colon-separated list _must_ equal the number of VCPUs of the
652 # Map the entire instance to CPUs 0-2
653 gnt-instance modify -H cpu_mask=0-2 my-inst
655 # Map vCPU 0 to physical CPU 1 and vCPU 1 to CPU 3 (assuming 2 vCPUs)
656 gnt-instance modify -H cpu_mask=1:3 my-inst
658 # Pin vCPU 0 to CPUs 1 or 2, and vCPU 1 to any CPU
659 gnt-instance modify -H cpu_mask=1-2:all my-inst
661 # Pin vCPU 0 to any CPU, vCPU 1 to CPUs 1, 3, 4 or 5, and CPU 2 to
662 # CPU 0 (backslashes for escaping the comma)
663 gnt-instance modify -H cpu_mask=all:1\\,3-5:0 my-inst
665 # Pin entire VM to CPU 0
666 gnt-instance modify -H cpu_mask=0 my-inst
668 # Turn off CPU pinning (default setting)
669 gnt-instance modify -H cpu_mask=all my-inst
672 Valid for the Xen hypervisor.
674 Set the maximum amount of cpu usage by the VM. The value is a percentage
675 between 0 and (100 * number of VCPUs). Default cap is 0: unlimited.
678 Valid for the Xen hypervisor.
680 Set the cpu time ratio to be allocated to the VM. Valid values are
681 between 1 and 65535. Default weight is 256.
684 Valid for the KVM hypervisor.
686 This option specifies the usb mouse type to be used. It can be
687 "mouse" or "tablet". When using VNC it's recommended to set it to
691 Valid for the KVM hypervisor.
693 This option specifies the keyboard mapping to be used. It is only
694 needed when using the VNC console. For example: "fr" or "en-gb".
697 Valid for Xen PVM, Xen HVM and KVM hypervisors.
699 Normally if an instance reboots, the hypervisor will restart it. If
700 this option is set to ``exit``, the hypervisor will treat a reboot
701 as a shutdown instead.
703 It is set to ``reboot`` by default.
706 Valid for the KVM hypervisor.
708 Number of emulated CPU cores.
711 Valid for the KVM hypervisor.
713 Number of emulated CPU threads.
716 Valid for the KVM hypervisor.
718 Number of emulated CPU sockets.
721 Valid for the KVM hypervisor.
723 Comma separated list of emulated sounds cards, or "all" to enable
724 all the available ones.
727 Valid for the KVM hypervisor.
729 Comma separated list of usb devices. These can be emulated devices
730 or passthrough ones, and each one gets passed to kvm with its own
731 ``-usbdevice`` option. See the **qemu**\(1) manpage for the syntax
732 of the possible components.
735 Valid for the KVM hypervisor.
737 Emulated vga mode, passed the the kvm -vga option.
740 Valid for the KVM hypervisor.
742 Any other option to the KVM hypervisor, useful tweaking anything
743 that Ganeti doesn't support. Note that values set with this
744 parameter are split on a space character and currently don't support
748 Valid for the KVM hypervisor.
750 Use in case an instance must be booted with an exact type of
751 machine version (due to e.g. outdated drivers). In case it's not set
752 the default version supported by your version of kvm is used.
755 Valid for the KVM hypervisor.
757 Path to the userspace KVM (or qemu) program.
760 Valid for the KVM hypervisor.
762 This boolean option determines whether the tap devices used by the
763 KVM paravirtual nics (virtio-net) will get created with VNET_HDR
764 (IFF_VNET_HDR) support.
766 If set to false, it effectively disables offloading on the virio-net
767 interfaces, which prevents host kernel tainting and log flooding,
768 when dealing with broken or malicious virtio-net drivers.
770 It is set to ``true`` by default.
772 The ``-O (--os-parameters)`` option allows customisation of the OS
773 parameters. The actual parameter names and values depends on the OS
774 being used, but the syntax is the same key=value. For example, setting
775 a hypothetical ``dhcp`` parameter to yes can be achieved by::
777 gnt-instance add -O dhcp=yes ...
779 The ``-I (--iallocator)`` option specifies the instance allocator plugin
780 to use (``.`` means the default allocator). If you pass in this option
781 the allocator will select nodes for this instance automatically, so you
782 don't need to pass them with the ``-n`` option. For more information
783 please refer to the instance allocator documentation.
785 The ``-t (--disk-template)`` options specifies the disk layout type
786 for the instance. If no disk template is specified, the default disk
787 template is used. The default disk template is the first in the list
788 of enabled disk templates, which can be adjusted cluster-wide with
789 ``gnt-cluster modify``. The available choices for disk templates are:
792 This creates an instance with no disks. Its useful for testing only
793 (or other special cases).
796 Disk devices will be regular files.
799 Disk devices will be regulare files on a shared directory.
802 Disk devices will be logical volumes.
805 Disk devices will be drbd (version 8.x) on top of lvm volumes.
808 Disk devices will be rbd volumes residing inside a RADOS cluster.
811 Disk devices will be adopted pre-existent block devices.
814 Disk devices will be provided by external shared storage,
815 through the ExtStorage Interface using ExtStorage providers.
817 The optional second value of the ``-n (--node)`` is used for the drbd
818 template type and specifies the remote node.
820 If you do not want gnt-instance to wait for the disk mirror to be
821 synced, use the ``--no-wait-for-sync`` option.
823 The ``--file-storage-dir`` specifies the relative path under the
824 cluster-wide file storage directory to store file-based disks. It is
825 useful for having different subdirectories for different
826 instances. The full path of the directory where the disk files are
827 stored will consist of cluster-wide file storage directory + optional
828 subdirectory + instance name. This option is only relevant for
829 instances using the file storage backend.
831 The ``--file-driver`` specifies the driver to use for file-based
832 disks. Note that currently these drivers work with the xen hypervisor
833 only. This option is only relevant for instances using the file
834 storage backend. The available choices are:
837 Kernel loopback driver. This driver uses loopback devices to
838 access the filesystem within the file. However, running I/O
839 intensive applications in your instance using the loop driver
840 might result in slowdowns. Furthermore, if you use the loopback
841 driver consider increasing the maximum amount of loopback devices
842 (on most systems it's 8) using the max\_loop param.
845 The blktap driver (for Xen hypervisors). In order to be able to
846 use the blktap driver you should check if the 'blktapctrl' user
847 space disk agent is running (usually automatically started via
848 xend). This user-level disk I/O interface has the advantage of
849 better performance. Especially if you use a network file system
850 (e.g. NFS) to store your instances this is the recommended choice.
852 If ``--ignore-ipolicy`` is given any instance policy violations occuring
853 during this operation are ignored.
855 See **ganeti**\(7) for a description of ``--submit`` and other common
860 # gnt-instance add -t file --disk 0:size=30g -B maxmem=512 -o debian-etch \
861 -n node1.example.com --file-storage-dir=mysubdir instance1.example.com
862 # gnt-instance add -t plain --disk 0:size=30g -B maxmem=1024,minmem=512 \
863 -o debian-etch -n node1.example.com instance1.example.com
864 # gnt-instance add -t plain --disk 0:size=30g --disk 1:size=100g,vg=san \
865 -B maxmem=512 -o debian-etch -n node1.example.com instance1.example.com
866 # gnt-instance add -t drbd --disk 0:size=30g -B maxmem=512 -o debian-etch \
867 -n node1.example.com:node2.example.com instance2.example.com
868 # gnt-instance add -t rbd --disk 0:size=30g -B maxmem=512 -o debian-etch \
869 -n node1.example.com instance1.example.com
870 # gnt-instance add -t ext --disk 0:size=30g,provider=pvdr1 -B maxmem=512 \
871 -o debian-etch -n node1.example.com instance1.example.com
872 # gnt-instance add -t ext --disk 0:size=30g,provider=pvdr1,param1=val1 \
873 --disk 1:size=40g,provider=pvdr2,param2=val2,param3=val3 -B maxmem=512 \
874 -o debian-etch -n node1.example.com instance1.example.com
881 | [{-I|\--iallocator} *instance allocator*]
882 | {instances\_file.json}
884 This command (similar to the Ganeti 1.2 **batcher** tool) submits
885 multiple instance creation jobs based on a definition file. This
886 file can contain all options which are valid when adding an instance
887 with the exception of the ``iallocator`` field. The IAllocator is,
888 for optimization purposes, only allowed to be set for the whole batch
889 operation using the ``--iallocator`` parameter.
891 The instance file must be a valid-formed JSON file, containing an
892 array of dictionaries with instance creation parameters. All parameters
893 (except ``iallocator``) which are valid for the instance creation
894 OP code are allowed. The most important ones are:
897 The FQDN of the new instance.
900 The disk template to use for the instance, the same as in the
904 Array of disk specifications. Each entry describes one disk as a
905 dictionary of disk parameters.
908 A dictionary of backend parameters.
911 The hypervisor for the instance.
914 A dictionary with the hypervisor options. If not passed, the default
915 hypervisor options will be inherited.
918 List of NICs that will be created for the instance. Each entry
919 should be a dict, with mac, ip, mode and link as possible keys.
920 Please don't provide the "mac, ip, mode, link" parent keys if you
921 use this method for specifying NICs.
924 The primary and optionally the secondary node to use for the
925 instance (in case an iallocator script is not used). If those
926 parameters are given, they have to be given consistently for all
927 instances in the batch operation.
930 whether to start the instance
933 Skip the check for already-in-use instance; see the description in
934 the **add** command for details.
937 Skip the name check for instances; see the description in the
938 **add** command for details.
940 file\_storage\_dir, file\_driver
941 Configuration for the file disk type, see the **add** command for
945 A simple definition for one instance can be (with most of the
946 parameters taken from the cluster defaults)::
951 "instance_name": "instance1.example.com",
952 "disk_template": "drbd",
953 "os_type": "debootstrap",
954 "disks": [{"size":"1024"}],
956 "hypervisor": "xen-pvm"
960 "instance_name": "instance2.example.com",
961 "disk_template": "drbd",
962 "os_type": "debootstrap",
963 "disks": [{"size":"4096", "mode": "rw", "vg": "xenvg"}],
965 "hypervisor": "xen-hvm",
966 "hvparams": {"acpi": true},
967 "beparams": {"maxmem": 512, "minmem": 256}
971 The command will display the job id for each submitted instance, as
974 # gnt-instance batch-create instances.json
975 Submitted jobs 37, 38
980 | **remove** [\--ignore-failures] [\--shutdown-timeout=*N*] [\--submit]
981 | [\--print-job-id] [\--force] {*instance*}
983 Remove an instance. This will remove all data from the instance and
984 there is *no way back*. If you are not sure if you use an instance
985 again, use **shutdown** first and leave it in the shutdown state for a
988 The ``--ignore-failures`` option will cause the removal to proceed
989 even in the presence of errors during the removal of the instance
990 (e.g. during the shutdown or the disk removal). If this option is not
991 given, the command will stop at the first error.
993 The ``--shutdown-timeout`` is used to specify how much time to wait
994 before forcing the shutdown (e.g. ``xm destroy`` in Xen, killing the
995 kvm process for KVM, etc.). By default two minutes are given to each
998 The ``--force`` option is used to skip the interactive confirmation.
1000 See **ganeti**\(7) for a description of ``--submit`` and other common
1005 # gnt-instance remove instance1.example.com
1012 | [\--no-headers] [\--separator=*SEPARATOR*] [\--units=*UNITS*] [-v]
1013 | [{-o|\--output} *[+]FIELD,...*] [\--filter] [instance...]
1015 Shows the currently configured instances with memory usage, disk
1016 usage, the node they are running on, and their run status.
1018 The ``--no-headers`` option will skip the initial header line. The
1019 ``--separator`` option takes an argument which denotes what will be
1020 used between the output fields. Both these options are to help
1023 The units used to display the numeric values in the output varies,
1024 depending on the options given. By default, the values will be
1025 formatted in the most appropriate unit. If the ``--separator`` option
1026 is given, then the values are shown in mebibytes to allow parsing by
1027 scripts. In both cases, the ``--units`` option can be used to enforce
1028 a given output unit.
1030 The ``-v`` option activates verbose mode, which changes the display of
1031 special field states (see **ganeti**\(7)).
1033 The ``-o (--output)`` option takes a comma-separated list of output
1034 fields. The available fields and their meaning are:
1036 @QUERY_FIELDS_INSTANCE@
1038 If the value of the option starts with the character ``+``, the new
1039 field(s) will be added to the default list. This allows one to quickly
1040 see the default list plus a few other fields, instead of retyping the
1041 entire list of fields.
1043 There is a subtle grouping about the available output fields: all
1044 fields except for ``oper_state``, ``oper_ram``, ``oper_vcpus`` and
1045 ``status`` are configuration value and not run-time values. So if you
1046 don't select any of the these fields, the query will be satisfied
1047 instantly from the cluster configuration, without having to ask the
1048 remote nodes for the data. This can be helpful for big clusters when
1049 you only want some data and it makes sense to specify a reduced set of
1052 If exactly one argument is given and it appears to be a query filter
1053 (see **ganeti**\(7)), the query result is filtered accordingly. For
1054 ambiguous cases (e.g. a single field name as a filter) the ``--filter``
1055 (``-F``) option forces the argument to be treated as a filter (e.g.
1056 ``gnt-instance list -F admin_state``).
1058 The default output field list is: ``name``, ``os``, ``pnode``,
1059 ``admin_state``, ``oper_state``, ``oper_ram``.
1065 **list-fields** [field...]
1067 Lists available fields for instances.
1073 **info** [-s \| \--static] [\--roman] {\--all \| *instance*}
1075 Show detailed information about the given instance(s). This is
1076 different from **list** as it shows detailed data about the instance's
1077 disks (especially useful for the drbd disk template).
1079 If the option ``-s`` is used, only information available in the
1080 configuration file is returned, without querying nodes, making the
1083 Use the ``--all`` to get info about all instances, rather than
1084 explicitly passing the ones you're interested in.
1086 The ``--roman`` option can be used to cause envy among people who like
1087 ancient cultures, but are stuck with non-latin-friendly cluster
1088 virtualization technologies.
1094 | [{-H|\--hypervisor-parameters} *HYPERVISOR\_PARAMETERS*]
1095 | [{-B|\--backend-parameters} *BACKEND\_PARAMETERS*]
1096 | [{-m|\--runtime-memory} *SIZE*]
1097 | [\--net add[:options...] \|
1098 | \--net [*N*:]add[,options...] \|
1099 | \--net [*ID*:]remove \|
1100 | \--net *ID*:modify[,options...]]
1101 | [\--disk add:size=*SIZE*[,options...] \|
1102 | \--disk *N*:add,size=*SIZE*[,options...] \|
1103 | \--disk *N*:add,size=*SIZE*,provider=*PROVIDER*[,options...][,param=*value*... ] \|
1104 | \--disk *ID*:modify[,options...]
1105 | \--disk [*ID*:]remove]
1106 | [{-t|\--disk-template} plain \| {-t|\--disk-template} drbd -n *new_secondary*] [\--no-wait-for-sync]
1107 | [\--new-primary=*node*]
1108 | [\--os-type=*OS* [\--force-variant]]
1109 | [{-O|\--os-parameters} *param*=*value*... ]
1110 | [\--offline \| \--online]
1111 | [\--submit] [\--print-job-id]
1112 | [\--ignore-ipolicy]
1115 Modifies the memory size, number of vcpus, ip address, MAC address
1116 and/or NIC parameters for an instance. It can also add and remove
1117 disks and NICs to/from the instance. Note that you need to give at
1118 least one of the arguments, otherwise the command complains.
1120 The ``-H (--hypervisor-parameters)``, ``-B (--backend-parameters)``
1121 and ``-O (--os-parameters)`` options specifies hypervisor, backend and
1122 OS parameter options in the form of name=value[,...]. For details
1123 which options can be specified, see the **add** command.
1125 The ``-t (--disk-template)`` option will change the disk template of
1126 the instance. Currently only conversions between the plain and drbd
1127 disk templates are supported, and the instance must be stopped before
1128 attempting the conversion. When changing from the plain to the drbd
1129 disk template, a new secondary node must be specified via the ``-n``
1130 option. The option ``--no-wait-for-sync`` can be used when converting
1131 to the ``drbd`` template in order to make the instance available for
1132 startup before DRBD has finished resyncing.
1134 The ``-m (--runtime-memory)`` option will change an instance's runtime
1135 memory to the given size (in MB if a different suffix is not specified),
1136 by ballooning it up or down to the new value.
1138 The ``--disk add:size=*SIZE*,[options..]`` option adds a disk to the
1139 instance, and ``--disk *N*:add:size=*SIZE*,[options..]`` will add a disk
1140 to the the instance at a specific index. The available options are the
1141 same as in the **add** command(``spindles``, ``mode``, ``name``, ``vg``,
1142 ``metavg``). When adding an ExtStorage disk the ``provider=*PROVIDER*``
1143 option is also mandatory and specifies the ExtStorage provider. Also,
1144 for ExtStorage disks arbitrary parameters can be passed as additional
1145 comma separated options, same as in the **add** command. -The ``--disk
1146 remove`` option will remove the last disk of the instance. Use ``--disk
1147 `` *ID*``:remove`` to remove a disk by its identifier. *ID* can be the
1148 index of the disk, the disks's name or the disks's UUID. The ``--disk
1149 *ID*:modify[,options...]`` will change the options of the disk.
1150 Available options are:
1153 The access mode. Either ``ro`` (read-only) or the default ``rw`` (read-write).
1156 This option specifies a name for the disk, which can be used as a disk
1157 identifier. An instance can not have two disks with the same name.
1159 The ``--net *N*:add[,options..]`` will add a new network interface to
1160 the instance. The available options are the same as in the **add**
1161 command (``mac``, ``ip``, ``link``, ``mode``, ``network``). The
1162 ``--net *ID*,remove`` will remove the intances' NIC with *ID* identifier,
1163 which can be the index of the NIC, the NIC's name or the NIC's UUID.
1164 The ``--net *ID*:modify[,options..]`` option will change the parameters of
1165 the instance network interface with the *ID* identifier.
1167 The option ``-o (--os-type)`` will change the OS name for the instance
1168 (without reinstallation). In case an OS variant is specified that is
1169 not found, then by default the modification is refused, unless
1170 ``--force-variant`` is passed. An invalid OS will also be refused,
1171 unless the ``--force`` option is given.
1173 The option ``--new-primary`` will set the new primary node of an instance
1174 assuming the disks have already been moved manually. Unless the ``--force``
1175 option is given, it is verified that the instance is no longer running
1176 on its current primary node.
1178 The ``--online`` and ``--offline`` options are used to transition an
1179 instance into and out of the ``offline`` state. An instance can be
1180 turned offline only if it was previously down. The ``--online`` option
1181 fails if the instance was not in the ``offline`` state, otherwise it
1182 changes instance's state to ``down``. These modifications take effect
1185 If ``--ignore-ipolicy`` is given any instance policy violations occuring
1186 during this operation are ignored.
1188 See **ganeti**\(7) for a description of ``--submit`` and other common
1191 Most of the changes take effect at the next restart. If the instance is
1192 running, there is no effect on the instance.
1197 | **reinstall** [{-o|\--os-type} *os-type*] [\--select-os] [-f *force*]
1198 | [\--force-multiple]
1199 | [\--instance \| \--node \| \--primary \| \--secondary \| \--all]
1200 | [{-O|\--os-parameters} *OS\_PARAMETERS*] [\--submit] [\--print-job-id]
1203 Reinstalls the operating system on the given instance(s). The
1204 instance(s) must be stopped when running this command. If the ``-o
1205 (--os-type)`` is specified, the operating system is changed.
1207 The ``--select-os`` option switches to an interactive OS reinstall.
1208 The user is prompted to select the OS template from the list of
1209 available OS templates. OS parameters can be overridden using ``-O
1210 (--os-parameters)`` (more documentation for this option under the
1213 Since this is a potentially dangerous command, the user will be
1214 required to confirm this action, unless the ``-f`` flag is passed.
1215 When multiple instances are selected (either by passing multiple
1216 arguments or by using the ``--node``, ``--primary``, ``--secondary``
1217 or ``--all`` options), the user must pass the ``--force-multiple``
1218 options to skip the interactive confirmation.
1220 See **ganeti**\(7) for a description of ``--submit`` and other common
1226 | **rename** [\--no-ip-check] [\--no-name-check] [\--submit] [\--print-job-id]
1227 | {*instance*} {*new\_name*}
1229 Renames the given instance. The instance must be stopped when running
1230 this command. The requirements for the new name are the same as for
1231 adding an instance: the new name must be resolvable and the IP it
1232 resolves to must not be reachable (in order to prevent duplicate IPs
1233 the next time the instance is started). The IP test can be skipped if
1234 the ``--no-ip-check`` option is passed.
1236 Note that you can rename an instance to its same name, to force
1237 re-executing the os-specific rename script for that instance, if
1240 The ``--no-name-check`` skips the check for the new instance name via
1241 the resolver (e.g. in DNS or /etc/hosts, depending on your setup) and
1242 that the resolved name matches the provided name. Since the name check
1243 is used to compute the IP address, if you pass this option you must also
1244 pass the ``--no-ip-check`` option.
1246 See **ganeti**\(7) for a description of ``--submit`` and other common
1249 Starting/stopping/connecting to console
1250 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1256 | [\--force] [\--ignore-offline]
1257 | [\--force-multiple] [\--no-remember]
1258 | [\--instance \| \--node \| \--primary \| \--secondary \| \--all \|
1259 | \--tags \| \--node-tags \| \--pri-node-tags \| \--sec-node-tags]
1260 | [{-H|\--hypervisor-parameters} ``key=value...``]
1261 | [{-B|\--backend-parameters} ``key=value...``]
1262 | [\--submit] [\--print-job-id] [\--paused]
1265 Starts one or more instances, depending on the following options. The
1266 four available modes are:
1269 will start the instances given as arguments (at least one argument
1270 required); this is the default selection
1273 will start the instances who have the given node as either primary
1277 will start all instances whose primary node is in the list of nodes
1278 passed as arguments (at least one node required)
1281 will start all instances whose secondary node is in the list of
1282 nodes passed as arguments (at least one node required)
1285 will start all instances in the cluster (no arguments accepted)
1288 will start all instances in the cluster with the tags given as
1292 will start all instances in the cluster on nodes with the tags
1296 will start all instances in the cluster on primary nodes with the
1297 tags given as arguments
1300 will start all instances in the cluster on secondary nodes with the
1301 tags given as arguments
1303 Note that although you can pass more than one selection option, the
1304 last one wins, so in order to guarantee the desired result, don't pass
1305 more than one such option.
1307 Use ``--force`` to start even if secondary disks are failing.
1308 ``--ignore-offline`` can be used to ignore offline primary nodes and
1309 mark the instance as started even if the primary is not available.
1311 The ``--force-multiple`` will skip the interactive confirmation in the
1312 case the more than one instance will be affected.
1314 The ``--no-remember`` option will perform the startup but not change
1315 the state of the instance in the configuration file (if it was stopped
1316 before, Ganeti will still think it needs to be stopped). This can be
1317 used for testing, or for a one shot-start where you don't want the
1318 watcher to restart the instance if it crashes.
1320 The ``-H (--hypervisor-parameters)`` and ``-B (--backend-parameters)``
1321 options specify temporary hypervisor and backend parameters that can
1322 be used to start an instance with modified parameters. They can be
1323 useful for quick testing without having to modify an instance back and
1326 # gnt-instance start -H kernel_args="single" instance1
1327 # gnt-instance start -B maxmem=2048 instance2
1330 The first form will start the instance instance1 in single-user mode,
1331 and the instance instance2 with 2GB of RAM (this time only, unless
1332 that is the actual instance memory size already). Note that the values
1333 override the instance parameters (and not extend them): an instance
1334 with "kernel\_args=ro" when started with -H kernel\_args=single will
1335 result in "single", not "ro single".
1337 The ``--paused`` option is only valid for Xen and kvm hypervisors. This
1338 pauses the instance at the start of bootup, awaiting ``gnt-instance
1339 console`` to unpause it, allowing the entire boot process to be
1340 monitored for debugging.
1342 See **ganeti**\(7) for a description of ``--submit`` and other common
1347 # gnt-instance start instance1.example.com
1348 # gnt-instance start --node node1.example.com node2.example.com
1349 # gnt-instance start --all
1357 | [\--force] [\--force-multiple] [\--ignore-offline] [\--no-remember]
1358 | [\--instance \| \--node \| \--primary \| \--secondary \| \--all \|
1359 | \--tags \| \--node-tags \| \--pri-node-tags \| \--sec-node-tags]
1360 | [\--submit] [\--print-job-id]
1363 Stops one or more instances. If the instance cannot be cleanly stopped
1364 during a hardcoded interval (currently 2 minutes), it will forcibly
1365 stop the instance (equivalent to switching off the power on a physical
1368 The ``--timeout`` is used to specify how much time to wait before
1369 forcing the shutdown (e.g. ``xm destroy`` in Xen, killing the kvm
1370 process for KVM, etc.). By default two minutes are given to each
1373 The ``--instance``, ``--node``, ``--primary``, ``--secondary``,
1374 ``--all``, ``--tags``, ``--node-tags``, ``--pri-node-tags`` and
1375 ``--sec-node-tags`` options are similar as for the **startup** command
1376 and they influence the actual instances being shutdown.
1378 ``--ignore-offline`` can be used to ignore offline primary nodes and
1379 force the instance to be marked as stopped. This option should be used
1380 with care as it can lead to an inconsistent cluster state.
1382 Use ``--force`` to be able to shutdown an instance even when it's marked
1383 as offline. This is useful is an offline instance ends up in the
1384 ``ERROR_up`` state, for example.
1386 The ``--no-remember`` option will perform the shutdown but not change
1387 the state of the instance in the configuration file (if it was running
1388 before, Ganeti will still thinks it needs to be running). This can be
1389 useful for a cluster-wide shutdown, where some instances are marked as
1390 up and some as down, and you don't want to change the running state:
1391 you just need to disable the watcher, shutdown all instances with
1392 ``--no-remember``, and when the watcher is activated again it will
1393 restore the correct runtime state for all instances.
1395 See **ganeti**\(7) for a description of ``--submit`` and other common
1400 # gnt-instance shutdown instance1.example.com
1401 # gnt-instance shutdown --all
1408 | [{-t|\--type} *REBOOT-TYPE*]
1409 | [\--ignore-secondaries]
1410 | [\--shutdown-timeout=*N*]
1411 | [\--force-multiple]
1412 | [\--instance \| \--node \| \--primary \| \--secondary \| \--all \|
1413 | \--tags \| \--node-tags \| \--pri-node-tags \| \--sec-node-tags]
1414 | [\--submit] [\--print-job-id]
1417 Reboots one or more instances. The type of reboot depends on the value
1418 of ``-t (--type)``. A soft reboot does a hypervisor reboot, a hard reboot
1419 does a instance stop, recreates the hypervisor config for the instance
1420 and starts the instance. A full reboot does the equivalent of
1421 **gnt-instance shutdown && gnt-instance startup**. The default is
1424 For the hard reboot the option ``--ignore-secondaries`` ignores errors
1425 for the secondary node while re-assembling the instance disks.
1427 The ``--instance``, ``--node``, ``--primary``, ``--secondary``,
1428 ``--all``, ``--tags``, ``--node-tags``, ``--pri-node-tags`` and
1429 ``--sec-node-tags`` options are similar as for the **startup** command
1430 and they influence the actual instances being rebooted.
1432 The ``--shutdown-timeout`` is used to specify how much time to wait
1433 before forcing the shutdown (xm destroy in xen, killing the kvm
1434 process, for kvm). By default two minutes are given to each instance
1437 The ``--force-multiple`` will skip the interactive confirmation in the
1438 case the more than one instance will be affected.
1440 See **ganeti**\(7) for a description of ``--submit`` and other common
1445 # gnt-instance reboot instance1.example.com
1446 # gnt-instance reboot --type=full instance1.example.com
1452 **console** [\--show-cmd] {*instance*}
1454 Connects to the console of the given instance. If the instance is not
1455 up, an error is returned. Use the ``--show-cmd`` option to display the
1456 command instead of executing it.
1458 For HVM instances, this will attempt to connect to the serial console
1459 of the instance. To connect to the virtualized "physical" console of a
1460 HVM instance, use a VNC client with the connection info from the
1463 For Xen/kvm instances, if the instance is paused, this attempts to
1464 unpause the instance after waiting a few seconds for the connection to
1465 the console to be made.
1469 # gnt-instance console instance1.example.com
1478 | **replace-disks** [\--submit] [\--print-job-id] [\--early-release]
1479 | [\--ignore-ipolicy] {-p} [\--disks *idx*] {*instance*}
1481 | **replace-disks** [\--submit] [\--print-job-id] [\--early-release]
1482 | [\--ignore-ipolicy] {-s} [\--disks *idx*] {*instance*}
1484 | **replace-disks** [\--submit] [\--print-job-id] [\--early-release]
1485 | [\--ignore-ipolicy]
1486 | {{-I\|\--iallocator} *name* \| {{-n|\--new-secondary} *node* } {*instance*}
1488 | **replace-disks** [\--submit] [\--print-job-id] [\--early-release]
1489 | [\--ignore-ipolicy] {-a\|\--auto} {*instance*}
1491 This command is a generalized form for replacing disks. It is
1492 currently only valid for the mirrored (DRBD) disk template.
1494 The first form (when passing the ``-p`` option) will replace the disks
1495 on the primary, while the second form (when passing the ``-s`` option
1496 will replace the disks on the secondary node. For these two cases (as
1497 the node doesn't change), it is possible to only run the replace for a
1498 subset of the disks, using the option ``--disks`` which takes a list
1499 of comma-delimited disk indices (zero-based), e.g. 0,2 to replace only
1500 the first and third disks.
1502 The third form (when passing either the ``--iallocator`` or the
1503 ``--new-secondary`` option) is designed to change secondary node of the
1504 instance. Specifying ``--iallocator`` makes the new secondary be
1505 selected automatically by the specified allocator plugin (use ``.`` to
1506 indicate the default allocator), otherwise the new secondary node will
1507 be the one chosen manually via the ``--new-secondary`` option.
1509 Note that it is not possible to select an offline or drained node as a
1512 The fourth form (when using ``--auto``) will automatically determine
1513 which disks of an instance are faulty and replace them within the same
1514 node. The ``--auto`` option works only when an instance has only
1515 faulty disks on either the primary or secondary node; it doesn't work
1516 when both sides have faulty disks.
1518 The ``--early-release`` changes the code so that the old storage on
1519 secondary node(s) is removed early (before the resync is completed)
1520 and the internal Ganeti locks for the current (and new, if any)
1521 secondary node are also released, thus allowing more parallelism in
1522 the cluster operation. This should be used only when recovering from a
1523 disk failure on the current secondary (thus the old storage is already
1524 broken) or when the storage on the primary node is known to be fine
1525 (thus we won't need the old storage for potential recovery).
1527 The ``--ignore-ipolicy`` let the command ignore instance policy
1528 violations if replace-disks changes groups and the instance would
1529 violate the new groups instance policy.
1531 See **ganeti**\(7) for a description of ``--submit`` and other common
1537 | **activate-disks** [\--submit] [\--print-job-id] [\--ignore-size]
1538 | [\--wait-for-sync] {*instance*}
1540 Activates the block devices of the given instance. If successful, the
1541 command will show the location and name of the block devices::
1543 node1.example.com:disk/0:/dev/drbd0
1544 node1.example.com:disk/1:/dev/drbd1
1547 In this example, *node1.example.com* is the name of the node on which
1548 the devices have been activated. The *disk/0* and *disk/1* are the
1549 Ganeti-names of the instance disks; how they are visible inside the
1550 instance is hypervisor-specific. */dev/drbd0* and */dev/drbd1* are the
1551 actual block devices as visible on the node.
1553 The ``--ignore-size`` option can be used to activate disks ignoring
1554 the currently configured size in Ganeti. This can be used in cases
1555 where the configuration has gotten out of sync with the real-world
1556 (e.g. after a partially-failed grow-disk operation or due to rounding
1557 in LVM devices). This should not be used in normal cases, but only
1558 when activate-disks fails without it.
1560 The ``--wait-for-sync`` option will ensure that the command returns only
1561 after the instance's disks are synchronised (mostly for DRBD); this can
1562 be useful to ensure consistency, as otherwise there are no commands that
1563 can wait until synchronisation is done. However when passing this
1564 option, the command will have additional output, making it harder to
1565 parse the disk information.
1567 Note that it is safe to run this command while the instance is already
1570 See **ganeti**\(7) for a description of ``--submit`` and other common
1576 **deactivate-disks** [-f] [\--submit] [\--print-job-id] {*instance*}
1578 De-activates the block devices of the given instance. Note that if you
1579 run this command for an instance with a drbd disk template, while it
1580 is running, it will not be able to shutdown the block devices on the
1581 primary node, but it will shutdown the block devices on the secondary
1582 nodes, thus breaking the replication.
1584 The ``-f``/``--force`` option will skip checks that the instance is
1585 down; in case the hypervisor is confused and we can't talk to it,
1586 normally Ganeti will refuse to deactivate the disks, but with this
1587 option passed it will skip this check and directly try to deactivate
1588 the disks. This can still fail due to the instance actually running or
1591 See **ganeti**\(7) for a description of ``--submit`` and other common
1597 | **grow-disk** [\--no-wait-for-sync] [\--submit] [\--print-job-id]
1599 | {*instance*} {*disk*} {*amount*}
1601 Grows an instance's disk. This is only possible for instances having a
1602 plain, drbd, file, sharedfile, rbd or ext disk template. For the ext
1603 template to work, the ExtStorage provider should also support growing.
1604 This means having a ``grow`` script that actually grows the volume of
1605 the external shared storage.
1607 Note that this command only change the block device size; it will not
1608 grow the actual filesystems, partitions, etc. that live on that
1609 disk. Usually, you will need to:
1611 #. use **gnt-instance grow-disk**
1613 #. reboot the instance (later, at a convenient time)
1615 #. use a filesystem resizer, such as **ext2online**\(8) or
1616 **xfs\_growfs**\(8) to resize the filesystem, or use **fdisk**\(8) to
1617 change the partition table on the disk
1619 The *disk* argument is the index of the instance disk to grow. The
1620 *amount* argument is given as a number which can have a suffix (like the
1621 disk size in instance create); if the suffix is missing, the value will
1622 be interpreted as mebibytes.
1624 By default, the *amount* value represents the desired increase in the
1625 disk size (e.g. an amount of 1G will take a disk of size 3G to 4G). If
1626 the optional ``--absolute`` parameter is passed, then the *amount*
1627 argument doesn't represent the delta, but instead the desired final disk
1628 size (e.g. an amount of 8G will take a disk of size 4G to 8G).
1630 For instances with a drbd template, note that the disk grow operation
1631 might complete on one node but fail on the other; this will leave the
1632 instance with different-sized LVs on the two nodes, but this will not
1633 create problems (except for unused space).
1635 If you do not want gnt-instance to wait for the new disk region to be
1636 synced, use the ``--no-wait-for-sync`` option.
1638 See **ganeti**\(7) for a description of ``--submit`` and other common
1641 Example (increase the first disk for instance1 by 16GiB)::
1643 # gnt-instance grow-disk instance1.example.com 0 16g
1645 Example for increasing the disk size to a certain size::
1647 # gnt-instance grow-disk --absolute instance1.example.com 0 32g
1649 Also note that disk shrinking is not supported; use **gnt-backup
1650 export** and then **gnt-backup import** to reduce the disk size of an
1656 | **recreate-disks** [\--submit] [\--print-job-id]
1657 | [{-n node1:[node2] \| {-I\|\--iallocator *name*}}]
1658 | [\--disk=*N*[:[size=*VAL*][,spindles=*VAL*][,mode=*ro\|rw*]]] {*instance*}
1660 Recreates all or a subset of disks of the given instance.
1662 Note that this functionality should only be used for missing disks; if
1663 any of the given disks already exists, the operation will fail. While
1664 this is suboptimal, recreate-disks should hopefully not be needed in
1665 normal operation and as such the impact of this is low.
1667 If only a subset should be recreated, any number of ``disk`` options can
1668 be specified. It expects a disk index and an optional list of disk
1669 parameters to change. Only ``size``, ``spindles``, and ``mode`` can be
1670 changed while recreating disks. To recreate all disks while changing
1671 parameters on a subset only, a ``--disk`` option must be given for every
1672 disk of the instance.
1674 Optionally the instance's disks can be recreated on different
1675 nodes. This can be useful if, for example, the original nodes of the
1676 instance have gone down (and are marked offline), so we can't recreate
1677 on the same nodes. To do this, pass the new node(s) via ``-n`` option,
1678 with a syntax similar to the **add** command. The number of nodes
1679 passed must equal the number of nodes that the instance currently
1680 has. Note that changing nodes is only allowed when all disks are
1681 replaced, e.g. when no ``--disk`` option is passed.
1683 Another method of choosing which nodes to place the instance on is by
1684 using the specified iallocator, passing the ``--iallocator`` option.
1685 The primary and secondary nodes will be chosen by the specified
1686 iallocator plugin, or by the default allocator if ``.`` is specified.
1688 See **ganeti**\(7) for a description of ``--submit`` and other common
1697 | **failover** [-f] [\--ignore-consistency] [\--ignore-ipolicy]
1698 | [\--shutdown-timeout=*N*]
1699 | [{-n|\--target-node} *node* \| {-I|\--iallocator} *name*]
1701 | [\--submit] [\--print-job-id]
1704 Failover will stop the instance (if running), change its primary node,
1705 and if it was originally running it will start it again (on the new
1706 primary). This works for instances with drbd template (in which case you
1707 can only fail to the secondary node) and for externally mirrored
1708 templates (sharedfile, blockdev, rbd and ext) (in which case you can
1709 fail to any other node).
1711 If the instance's disk template is of type sharedfile, blockdev, rbd or
1712 ext, then you can explicitly specify the target node (which can be any
1713 node) using the ``-n`` or ``--target-node`` option, or specify an
1714 iallocator plugin using the ``-I`` or ``--iallocator`` option. If you
1715 omit both, the default iallocator will be used to specify the target
1718 If the instance's disk template is of type drbd, the target node is
1719 automatically selected as the drbd's secondary node. Changing the
1720 secondary node is possible with a replace-disks operation.
1722 Normally the failover will check the consistency of the disks before
1723 failing over the instance. If you are trying to migrate instances off
1724 a dead node, this will fail. Use the ``--ignore-consistency`` option
1725 for this purpose. Note that this option can be dangerous as errors in
1726 shutting down the instance will be ignored, resulting in possibly
1727 having the instance running on two machines in parallel (on
1728 disconnected DRBD drives).
1730 The ``--shutdown-timeout`` is used to specify how much time to wait
1731 before forcing the shutdown (xm destroy in xen, killing the kvm
1732 process, for kvm). By default two minutes are given to each instance
1735 If ``--ignore-ipolicy`` is given any instance policy violations occuring
1736 during this operation are ignored.
1738 If the ``--cleanup`` option is passed, the operation changes from
1739 performin a failover to attempting recovery from a failed previous failover.
1740 In this mode, Ganeti checks if the instance runs on the correct node (and
1741 updates its configuration if not) and ensures the instances' disks
1742 are configured correctly.
1744 See **ganeti**\(7) for a description of ``--submit`` and other common
1749 # gnt-instance failover instance1.example.com
1751 For externally mirrored templates also ``-n`` is available::
1753 # gnt-instance failover -n node3.example.com instance1.example.com
1759 | **migrate** [-f] [\--allow-failover] [\--non-live]
1760 | [\--migration-mode=live\|non-live] [\--ignore-ipolicy]
1761 | [\--no-runtime-changes] [\--submit] [\--print-job-id]
1762 | [{-n|\--target-node} *node* \| {-I|\--iallocator} *name*] {*instance*}
1764 | **migrate** [-f] \--cleanup [\--submit] [\--print-job-id] {*instance*}
1766 Migrate will move the instance to its secondary node without shutdown.
1767 As with failover, it works for instances having the drbd disk template
1768 or an externally mirrored disk template type such as sharedfile,
1769 blockdev, rbd or ext.
1771 If the instance's disk template is of type sharedfile, blockdev, rbd or
1772 ext, then you can explicitly specify the target node (which can be any
1773 node) using the ``-n`` or ``--target-node`` option, or specify an
1774 iallocator plugin using the ``-I`` or ``--iallocator`` option. If you
1775 omit both, the default iallocator will be used to specify the target
1776 node. Alternatively, the default iallocator can be requested by
1777 specifying ``.`` as the name of the plugin.
1779 If the instance's disk template is of type drbd, the target node is
1780 automatically selected as the drbd's secondary node. Changing the
1781 secondary node is possible with a replace-disks operation.
1783 The migration command needs a perfectly healthy instance for drbd
1784 instances, as we rely on the dual-master capability of drbd8 and the
1785 disks of the instance are not allowed to be degraded.
1787 The ``--non-live`` and ``--migration-mode=non-live`` options will
1788 switch (for the hypervisors that support it) between a "fully live"
1789 (i.e. the interruption is as minimal as possible) migration and one in
1790 which the instance is frozen, its state saved and transported to the
1791 remote node, and then resumed there. This all depends on the
1792 hypervisor support for two different methods. In any case, it is not
1793 an error to pass this parameter (it will just be ignored if the
1794 hypervisor doesn't support it). The option ``--migration-mode=live``
1795 option will request a fully-live migration. The default, when neither
1796 option is passed, depends on the hypervisor parameters (and can be
1797 viewed with the **gnt-cluster info** command).
1799 If the ``--cleanup`` option is passed, the operation changes from
1800 migration to attempting recovery from a failed previous migration. In
1801 this mode, Ganeti checks if the instance runs on the correct node (and
1802 updates its configuration if not) and ensures the instances' disks
1803 are configured correctly. In this mode, the ``--non-live`` option is
1806 The option ``-f`` will skip the prompting for confirmation.
1808 If ``--allow-failover`` is specified it tries to fallback to failover if
1809 it already can determine that a migration won't work (e.g. if the
1810 instance is shut down). Please note that the fallback will not happen
1811 during execution. If a migration fails during execution it still fails.
1813 If ``--ignore-ipolicy`` is given any instance policy violations occuring
1814 during this operation are ignored.
1816 The ``--no-runtime-changes`` option forbids migrate to alter an
1817 instance's runtime before migrating it (eg. ballooning an instance
1818 down because the target node doesn't have enough available memory).
1820 If an instance has the backend parameter ``always_failover`` set to
1821 true, then the migration is automatically converted into a failover.
1823 See **ganeti**\(7) for a description of ``--submit`` and other common
1826 Example (and expected output)::
1828 # gnt-instance migrate instance1
1829 Instance instance1 will be migrated. Note that migration
1830 might impact the instance if anything goes wrong (e.g. due to bugs in
1831 the hypervisor). Continue?
1833 Migrating instance instance1.example.com
1834 * checking disk consistency between source and target
1835 * switching node node2.example.com to secondary mode
1836 * changing into standalone mode
1837 * changing disks into dual-master mode
1838 * wait until resync is done
1839 * preparing node2.example.com to accept the instance
1840 * migrating instance to node2.example.com
1841 * switching node node1.example.com to secondary mode
1842 * wait until resync is done
1843 * changing into standalone mode
1844 * changing disks into single-master mode
1845 * wait until resync is done
1853 | **move** [-f] [\--ignore-consistency]
1854 | [-n *node*] [\--shutdown-timeout=*N*] [\--submit] [\--print-job-id]
1855 | [\--ignore-ipolicy]
1858 Move will move the instance to an arbitrary node in the cluster. This
1859 works only for instances having a plain or file disk template.
1861 Note that since this operation is done via data copy, it will take a
1862 long time for big disks (similar to replace-disks for a drbd
1865 The ``--shutdown-timeout`` is used to specify how much time to wait
1866 before forcing the shutdown (e.g. ``xm destroy`` in XEN, killing the
1867 kvm process for KVM, etc.). By default two minutes are given to each
1870 The ``--ignore-consistency`` option will make Ganeti ignore any errors
1871 in trying to shutdown the instance on its node; useful if the
1872 hypervisor is broken and you want to recover the data.
1874 If ``--ignore-ipolicy`` is given any instance policy violations occuring
1875 during this operation are ignored.
1877 See **ganeti**\(7) for a description of ``--submit`` and other common
1882 # gnt-instance move -n node3.example.com instance1.example.com
1888 | **change-group** [\--submit] [\--print-job-id]
1889 | [\--iallocator *NAME*] [\--to *GROUP*...] {*instance*}
1891 This command moves an instance to another node group. The move is
1892 calculated by an iallocator, either given on the command line or as a
1895 If no specific destination groups are specified using ``--to``, all
1896 groups except the one containing the instance are considered.
1898 See **ganeti**\(7) for a description of ``--submit`` and other common
1903 # gnt-instance change-group -I hail --to rack2 inst1.example.com
1912 **add-tags** [\--from *file*] {*instancename*} {*tag*...}
1914 Add tags to the given instance. If any of the tags contains invalid
1915 characters, the entire operation will abort.
1917 If the ``--from`` option is given, the list of tags will be extended
1918 with the contents of that file (each line becomes a tag). In this
1919 case, there is not need to pass tags on the command line (if you do,
1920 both sources will be used). A file name of ``-`` will be interpreted
1926 **list-tags** {*instancename*}
1928 List the tags of the given instance.
1933 **remove-tags** [\--from *file*] {*instancename*} {*tag*...}
1935 Remove tags from the given instance. If any of the tags are not
1936 existing on the node, the entire operation will abort.
1938 If the ``--from`` option is given, the list of tags to be removed will
1939 be extended with the contents of that file (each line becomes a tag).
1940 In this case, there is not need to pass tags on the command line (if
1941 you do, tags from both sources will be removed). A file name of ``-``
1942 will be interpreted as stdin.
1944 .. vim: set textwidth=72 :