1 <!doctype refentry PUBLIC "-//OASIS//DTD DocBook V4.1//EN" [
3 <!-- Fill in your name for FIRSTNAME and SURNAME. -->
4 <!-- Please adjust the date whenever revising the manpage. -->
5 <!ENTITY dhdate "<date>February 11, 2009</date>">
6 <!-- SECTION should be 1-8, maybe w/ subsection other parameters are
7 allowed: see man(7), man(1). -->
8 <!ENTITY dhsection "<manvolnum>8</manvolnum>">
9 <!ENTITY dhucpackage "<refentrytitle>gnt-instance</refentrytitle>">
10 <!ENTITY dhpackage "gnt-instance">
12 <!ENTITY debian "<productname>Debian</productname>">
13 <!ENTITY gnu "<acronym>GNU</acronym>">
14 <!ENTITY gpl "&gnu; <acronym>GPL</acronym>">
15 <!ENTITY footer SYSTEM "footer.sgml">
25 <holder>Google Inc.</holder>
33 <refmiscinfo>ganeti 2.0</refmiscinfo>
36 <refname>&dhpackage;</refname>
38 <refpurpose>ganeti instance administration</refpurpose>
42 <command>&dhpackage; </command>
44 <arg choice="req">command</arg>
45 <arg>arguments...</arg>
49 <title>DESCRIPTION</title>
52 The <command>&dhpackage;</command> is used for instance
53 administration in the ganeti system.
58 <title>COMMANDS</title>
61 <title>Creation/removal/querying</title>
66 <command>add</command>
68 <arg choice="req">-t<group choice="req">
77 <arg rep="repeat">--disk=<replaceable>N</replaceable>:size=<replaceable>VAL</replaceable><arg>,mode=<replaceable>ro|rw</replaceable></arg></arg>
78 <arg>-s <replaceable>SIZE</replaceable></arg>
82 <arg rep="repeat">--net=<replaceable>N</replaceable><arg rep="repeat">:options</arg></arg>
86 <arg>-B <replaceable>BEPARAMS</replaceable></arg>
89 <arg>-H <replaceable>HYPERVISOR</replaceable><arg>:<arg choice="plain" rep="repeat">option=<replaceable>value</replaceable></arg></arg></arg>
92 <arg>--file-storage-dir <replaceable>dir_path</replaceable></arg>
93 <arg>--file-driver<group choice="req">
100 <arg>-n <replaceable>node<optional>:secondary-node</optional></replaceable></arg>
101 <arg>--iallocator <replaceable>name</replaceable></arg>
105 <arg choice="req">-o <replaceable>os-type</replaceable></arg>
110 <arg choice="req"><replaceable>instance</replaceable></arg>
114 Creates a new instance on the specified host. The
115 <replaceable>instance</replaceable> argument must be in DNS,
116 but depending on the bridge setup, need not be in the same
117 network as the nodes in the cluster.
121 The <option>disk</option> option specifies the parameters
122 for the disks of the instance. The numbering of disks starts
123 at zero, and at least one disk needs to be passed. For each
124 disk, at least the size needs to be given, and optionally
125 the access mode (read-only or the default of read-write) can
126 also be specified. The size is interpreted (when no unit is
127 given) in mebibytes. You can also use one of the suffixes
128 <literal>m</literal>, <literal>g</literal> or
129 <literal>t</literal> to specificy the exact the units used;
130 these suffixes map to mebibytes, gibibytes and tebibytes.
134 Alternatively, a single-disk instance can be created via the
135 <option>-s</option> option which takes a single argument,
136 the size of the disk. This is similar to the Ganeti 1.2
137 version (but will only create one disk).
141 The minimum disk specification is therefore
142 <userinput>--disk 0:size=20G</userinput> (or <userinput>-s
143 20G</userinput> when using the <option>-s</option> option),
144 and a three-disk instance can be specified as
145 <userinput>--disk 0:size=20G --disk 1:size=4G --disk
146 2:size=100G</userinput>.
150 The NICs of the instances can be specified via the
151 <option>--nic</option> option. By default, one NIC is
152 created for the instance, with a random MAC, and connected
153 to the default bridge. Each NIC can take up to three
154 parameters (all optional):
159 <simpara>either a value or <constant>GENERATE</constant>
160 to generate a new unique MAC</simpara>
166 <simpara>specifies the IP address assigned to the
167 instance from the Ganeti side (this is not necessarily
168 what the instance will use, but what the node expects
169 the instance to use)</simpara>
175 <simpara>specifies the bridge to attach this NIC
183 Alternatively, if no network is desired for the instance, you
184 can prevent the default of one NIC with the
185 <option>--no-nics</option> option.
189 The <option>-o</option> options specifies the operating
190 system to be installed. The available operating systems can
191 be listed with <command>gnt-os list</command>.
195 The <option>-B</option> option specifies the backend
196 parameters for the instance. If no such parameters are
197 specified, the values are inherited from the cluster. Possible
203 <simpara>the memory size of the instance; as usual,
204 suffixes can be used to denote the unit, otherwise the
205 value is taken in mebibites</simpara>
211 <simpara>the number of VCPUs to assign to the instance
212 (if this value makes sense for the hypervisor)</simpara>
216 <term>auto_balance</term>
218 <simpara>whether the instance is considered in the N+1
219 cluster checks (enough redundancy in the cluster to
220 survive a node failure)</simpara>
227 The <option>-H</option> option specified the hypervisor to
228 use for the instance (must be one of the enabled hypervisors
229 on the cluster) and optionally custom parameters for this
230 instance. If not other options are used (i.e. the invocation
231 is just <userinput>-H
232 <replaceable>NAME</replaceable></userinput>) the instance
233 will inherit the cluster options. The defaults below show
234 the cluster defaults at cluster creation time.
238 The possible hypervisor options are as follows:
241 <term>boot_order</term>
243 <simpara>Valid for the Xen HVM and KVM
244 hypervisors.</simpara>
246 <simpara>A string value denoting the boot order. This
247 has different meaning for the Xen HVM hypervisor and
248 for the KVM one.</simpara>
251 For Xen HVM, The boot order is a string of letters
252 listing the boot devices, with valid device letters
290 The default is not to set an HVM boot order which is
297 <term>cdrom_image_path</term>
299 <simpara>Valid for the Xen HVM and KVM hypervisors.</simpara>
301 <simpara>The path to a CDROM image to attach to the
307 <term>nic_type</term>
309 <simpara>Valid for the Xen HVM and KVM hypervisors.</simpara>
312 This parameter determines the way the network cards
313 are presented to the instance. The possible options are:
315 <member>rtl8139 (default for Xen HVM) (HVM & KVM)</member>
316 <member>ne2k_isa (HVM & KVM)</member>
317 <member>ne2k_pci (HVM & KVM)</member>
318 <member>i82551 (KVM)</member>
319 <member>i82557b (KVM)</member>
320 <member>i82559er (KVM)</member>
321 <member>pcnet (KVM)</member>
322 <member>e1000 (KVM)</member>
323 <member>paravirtual (default for KVM) (HVM & KVM)</member>
329 <term>disk_type</term>
331 <simpara>Valid for the Xen HVM and KVM hypervisors.</simpara>
334 This parameter determines the way the disks are
335 presented to the instance. The possible options are:
337 <member>ioemu (default for HVM & KVM) (HVM & KVM)</member>
338 <member>ide (HVM & KVM)</member>
339 <member>scsi (KVM)</member>
340 <member>sd (KVM)</member>
341 <member>mtd (KVM)</member>
342 <member>pflash (KVM)</member>
348 <term>vnc_bind_address</term>
350 <simpara>Valid for the Xen HVM and KVM hypervisors.</simpara>
352 <para>Specifies the address that the VNC listener for
353 this instance should bind to. Valid values are IPv4
354 addresses. Use the address 0.0.0.0 to bind to all
355 available interfaces (this is the default) or specify
356 the address of one of the interfaces on the node to
357 restrict listening to that interface.</para>
364 <simpara>Valid for the KVM hypervisor.</simpara>
366 <simpara>A boolean option that controls whether the
367 VNC connection is secured with TLS.</simpara>
372 <term>vnc_x509_path</term>
374 <simpara>Valid for the KVM hypervisor.</simpara>
376 <para>If <option>vnc_tls</option> is enabled, this
377 options specifies the path to the x509 certificate to
383 <term>vnc_x509_verify</term>
385 <simpara>Valid for the KVM hypervisor.</simpara>
392 <simpara>Valid for the Xen HVM and KVM hypervisors.</simpara>
395 A boolean option that specifies if the hypervisor
396 should enable ACPI support for this instance. By
397 default, ACPI is disabled.
405 <simpara>Valid for the Xen HVM and KVM hypervisors.</simpara>
408 A boolean option that specifies if the hypervisor
409 should enabled PAE support for this instance. The
410 default is false, disabling PAE support.
416 <term>kernel_path</term>
418 <simpara>Valid for the Xen PVM and KVM hypervisors.</simpara>
421 This option specifies the path (on the node) to the
422 kernel to boot the instance with. Xen PVM instances
423 always require this, while for KVM if this option is
424 empty, it will cause the machine to load the kernel
431 <term>kernel_args</term>
433 <simpara>Valid for the Xen PVM and KVM hypervisors.</simpara>
436 This options specifies extra arguments to the kernel
437 that will be loaded. device. This is always used
438 for Xen PVM, while for KVM it is only used if the
439 <option>kernel_path</option> option is also
444 The default setting for this value is simply
445 <constant>"ro"</constant>, which mounts the root
446 disk (initially) in read-only one. For example,
447 setting this to <userinput>single</userinput> will
448 cause the instance to start in single-user mode.
454 <term>initrd_path</term>
456 <simpara>Valid for the Xen PVM and KVM hypervisors.</simpara>
459 This option specifies the path (on the node) to the
460 initrd to boot the instance with. Xen PVM instances
461 can use this always, while for KVM if this option is
462 only used if the <option>kernel_path</option> option
469 <term>root_path</term>
471 <simpara>Valid for the Xen PVM and KVM hypervisors.</simpara>
474 This options specifies the name of the root
475 device. This is always needed for Xen PVM, while for
476 KVM it is only used if the
477 <option>kernel_path</option> option is also
484 <term>serial_console</term>
486 <simpara>Valid for the KVM hypervisor.</simpara>
488 <simpara>This boolean option specifies whether to
489 emulate a serial console for the instance.</simpara>
499 The <option>--iallocator</option> option specifies the instance
500 allocator plugin to use. If you pass in this option the allocator
501 will select nodes for this instance automatically, so you don't need
502 to pass them with the <option>-n</option> option. For more
503 information please refer to the instance allocator documentation.
507 The <option>-t</option> options specifies the disk layout type for
508 the instance. The available choices are:
511 <term>diskless</term>
514 This creates an instance with no disks. Its useful for
515 testing only (or other special cases).
522 <para>Disk devices will be regular files.</para>
528 <para>Disk devices will be logical volumes.</para>
535 Disk devices will be drbd (version 8.x) on top of
544 The optional second value of the <option>--node</option> is used for
545 the drbd template type and specifies the remote node.
549 If you do not want gnt-instance to wait for the disk mirror
550 to be synced, use the <option>--no-wait-for-sync</option>
555 The <option>--file-storage-dir</option> specifies the relative path
556 under the cluster-wide file storage directory to store file-based
557 disks. It is useful for having different subdirectories for
558 different instances. The full path of the directory where the disk
559 files are stored will consist of cluster-wide file storage directory
560 + optional subdirectory + instance name. Example:
561 /srv/ganeti/file-storage/mysubdir/instance1.example.com. This option
562 is only relevant for instances using the file storage backend.
566 The <option>--file-driver</option> specifies the driver to use for
567 file-based disks. Note that currently these drivers work with the
568 xen hypervisor only. This option is only relevant for instances using
569 the file storage backend. The available choices are:
575 Kernel loopback driver. This driver uses loopback
576 devices to access the filesystem within the
577 file. However, running I/O intensive applications in
578 your instance using the loop driver might result in
579 slowdowns. Furthermore, if you use the loopback
580 driver consider increasing the maximum amount of
581 loopback devices (on most systems it's 8) using the
589 <para>The blktap driver (for Xen hypervisors). In
590 order to be able to use the blktap driver you should
591 check if the 'blktapctrl' user space disk agent is
592 running (usually automatically started via xend). This
593 user-level disk I/O interface has the advantage of
594 better performance. Especially if you use a network
595 file system (e.g. NFS) to store your instances this is
596 the recommended choice.
604 The <option>--submit</option> option is used to send the job to
605 the master daemon but not wait for its completion. The job
606 ID will be shown so that it can be examined via
607 <command>gnt-job info</command>.
613 # gnt-instance add -t file --disk 0:size=30g -B memory=512 -o debian-etch \
614 -n node1.example.com --file-storage-dir=mysubdir instance1.example.com
615 # gnt-instance add -t plain --disk 0:size=30g -B memory=512 -o debian-etch \
616 -n node1.example.com instance1.example.com
617 # gnt-instance add -t drbd --disk 0:size=30g -B memory=512 -o debian-etch \
618 -n node1.example.com:node2.example.com instance2.example.com
624 <title>BATCH-CREATE</title>
626 <command>batch-create</command>
627 <arg choice="req">instances_file.json</arg>
631 This command (similar to the Ganeti 1.2
632 <command>batcher</command> tool) submits multiple instance
633 creation jobs based on a definition file. The instance
634 configurations do not encompass all the possible options for
635 the <command>add</command> command, but only a subset.
639 The instance file should be a valid-formed JSON file,
640 containing a dictionary with instance name and instance
641 parameters. The accepted parameters are:
645 <term>disk_size</term>
647 <simpara>The size of the disks of the instance.</simpara>
651 <term>disk_templace</term>
653 <simpara>The disk template to use for the instance,
654 the same as in the <command>add</command>
661 <simpara>A dictionary of backend parameters.</simpara>
665 <term>hypervisor</term>
667 <simpara>A dictionary with a single key (the
668 hypervisor name), and as value the hypervisor
669 options. If not passed, the default hypervisor and
670 hypervisor options will be inherited.</simpara>
674 <term>mac, ip, bridge</term>
676 <simpara>Specifications for the one NIC that will be
677 created for the instance.</simpara>
681 <term>primary_node, secondary_node</term>
683 <simpara>The primary and optionally the secondary node
684 to use for the instance (in case an iallocator script
685 is not used).</simpara>
689 <term>iallocator</term>
691 <simpara>Instead of specifying the nodes, an
692 iallocator script can be used to automatically compute
699 <simpara>whether to start the instance</simpara>
703 <term>ip_check</term>
705 <simpara>Skip the check for already-in-use instance;
706 see the description in the <command>add</command>
707 command for details.</simpara>
711 <term>file_storage_dir, file_driver</term>
713 <simpara>Configuration for the <literal>file</literal>
714 disk type, see the <command>add</command> command for
722 A simple definition for one instance can be (with most of
723 the parameters taken from the cluster defaults):
729 "disk_size": ["25G"],
735 "disk_size": ["25G"],
736 "iallocator": "dumb",
737 "hypervisor": "xen-hvm",
738 "hvparams": {"acpi": true},
739 "backend": {"memory": 512}
746 The command will display the job id for each submitted instance, as follows:
748 # gnt-instance batch-create instances.json
757 <title>REMOVE</title>
760 <command>remove</command>
761 <arg>--ignore-failures</arg>
763 <arg choice="req"><replaceable>instance</replaceable></arg>
767 Remove an instance. This will remove all data from the
768 instance and there is <emphasis>no way back</emphasis>. If
769 you are not sure if you use an instance again, use
770 <command>shutdown</command> first and leave it in the
771 shutdown state for a while.
776 The <option>--ignore-failures</option> option will cause the
777 removal to proceed even in the presence of errors during the
778 removal of the instance (e.g. during the shutdown or the
779 disk removal). If this option is not given, the command will
780 stop at the first error.
784 The <option>--submit</option> option is used to send the job to
785 the master daemon but not wait for its completion. The job
786 ID will be shown so that it can be examined via
787 <command>gnt-job info</command>.
793 # gnt-instance remove instance1.example.com
802 <command>list</command>
803 <arg>--no-headers</arg>
804 <arg>--separator=<replaceable>SEPARATOR</replaceable></arg>
805 <arg>-o <replaceable>[+]FIELD,...</replaceable></arg>
806 <arg rep="repeat">instance</arg>
810 Shows the currently configured instances with memory usage,
811 disk usage, the node they are running on, and their run
816 The <option>--no-headers</option> option will skip the
817 initial header line. The <option>--separator</option> option
818 takes an argument which denotes what will be used between
819 the output fields. Both these options are to help scripting.
823 The <option>-o</option> option takes a comma-separated list
824 of output fields. The available fields and their meaning
830 <simpara>the instance name</simpara>
836 <simpara>the OS of the instance</simpara>
842 <simpara>the primary node of the instance</simpara>
848 <simpara>comma-separated list of secondary nodes for the
849 instance; usually this will be just one node</simpara>
853 <term>admin_state</term>
855 <simpara>the desired state of the instance (either "yes"
856 or "no" denoting the instance should run or
861 <term>disk_template</term>
863 <simpara>the disk template of the instance</simpara>
867 <term>oper_state</term>
869 <simpara>the actual state of the instance; can be
870 one of the values "running", "stopped", "(node
877 <simpara>combined form of admin_state and oper_stat;
879 <computeroutput>ERROR_nodedown</computeroutput> if the
880 node of the instance is down,
881 <computeroutput>ERROR_down</computeroutput> if the
882 instance should run but is down,
883 <computeroutput>ERROR_up</computeroutput> if the
884 instance should be stopped but is actually running,
885 <computeroutput>ADMIN_down</computeroutput> if the
886 instance has been stopped (and is stopped) and
887 <computeroutput>running</computeroutput> if the
888 instance is set to be running (and is
893 <term>oper_ram</term>
895 <simpara>the actual memory usage of the instance as seen
896 by the hypervisor</simpara>
902 <simpara>the ip address ganeti recognizes as associated with
903 the first instance interface</simpara>
909 <simpara>the first instance interface MAC address</simpara>
915 <simpara>the bridge of the first instance NIC
920 <term>sda_size</term>
922 <simpara>the size of the instance's first disk</simpara>
926 <term>sdb_size</term>
928 <simpara>the size of the instance's second disk, if
935 <simpara>the number of VCPUs allocated to the
942 <simpara>comma-separated list of the instances's
947 <term>serial_no</term>
949 <simpara>the so called 'serial number' of the
950 instance; this is a numeric field that is incremented
951 each time the instance is modified, and it can be used
952 to track modifications</simpara>
956 <term>network_port</term>
958 <simpara>If the instance has a network port assigned
959 to it (e.g. for VNC connections), this will be shown,
960 otherwise <literal>-</literal> will be
965 <term>beparams</term>
967 <simpara>A text format of the entire beparams for the
968 instance. It's more useful to select individual fields
969 from this dictionary, see below.</simpara>
973 <term>disk.count</term>
975 <simpara>The number of instance disks.</simpara>
979 <term>disk.size/N</term>
981 <simpara>The size of the instance's Nth disk. This is
982 a more generic form of the <literal>sda_size</literal>
983 and <literal>sdb_size</literal> fields.</simpara>
987 <term>disk.sizes</term>
989 <simpara>A comma-separated list of the disk sizes for
990 this instance.</simpara>
994 <term>disk_usage</term>
996 <simpara>The total disk space used by this instance on
997 each of its nodes. This is not the instance-visible
998 disk size, but the actual disk "cost" of the
1003 <term>nic.mac/N</term>
1005 <simpara>The MAC of the Nth instance NIC.</simpara>
1009 <term>nic.ip/N</term>
1011 <simpara>The IP address of the Nth instance NIC.</simpara>
1015 <term>nic.bridge/N</term>
1017 <simpara>The bridge the Nth instance NIC is attached
1022 <term>nic.macs</term>
1024 <simpara>A comma-separated list of all the MACs of the
1025 instance's NICs.</simpara>
1029 <term>nic.ips</term>
1031 <simpara>A comma-separated list of all the IP
1032 addresses of the instance's NICs.</simpara>
1036 <term>nic.bridges</term>
1038 <simpara>A comma-separated list of all the bridges of the
1039 instance's NICs.</simpara>
1043 <term>nic.count</term>
1045 <simpara>The number of instance nics.</simpara>
1049 <term>hv/<replaceable>NAME</replaceable></term>
1051 <simpara>The value of the hypervisor parameter called
1052 <replaceable>NAME</replaceable>. For details of what
1053 hypervisor parameters exist and their meaning, see the
1054 <command>add</command> command.</simpara>
1058 <term>be/memory</term>
1060 <simpara>The configured memory for the instance.</simpara>
1064 <term>be/vcpus</term>
1066 <simpara>The configured number of VCPUs for the
1071 <term>be/auto_balance</term>
1073 <simpara>Whether the instance is considered in N+1
1081 If the value of the option starts with the character
1082 <constant>+</constant>, the new field(s) will be added to the
1083 default list. This allows to quickly see the default list
1084 plus a few other fields, instead of retyping the entire list
1089 There is a subtle grouping about the available output
1090 fields: all fields except for <option>oper_state</option>,
1091 <option>oper_ram</option> and <option>status</option> are
1092 configuration value and not run-time values. So if you don't
1093 select any of the these fields, the query will be satisfied
1094 instantly from the cluster configuration, without having to
1095 ask the remote nodes for the data. This can be helpful for
1096 big clusters when you only want some data and it makes sense
1097 to specify a reduced set of output fields.
1100 <para>The default output field list is:
1101 <simplelist type="inline">
1102 <member>name</member>
1104 <member>pnode</member>
1105 <member>admin_state</member>
1106 <member>oper_state</member>
1107 <member>oper_ram</member>
1116 <command>info</command>
1121 <group choice="req">
1123 <arg rep="repeat"><replaceable>instance</replaceable></arg>
1128 Show detailed information about the given instance(s). This is
1129 different from <command>list</command> as it shows detailed data
1130 about the instance's disks (especially useful for the drbd disk
1135 If the option <option>-s</option> is used, only information
1136 available in the configuration file is returned, without
1137 querying nodes, making the operation faster.
1141 Use the <option>--all</option> to get info about all instances,
1142 rather than explicitely passing the ones you're interested in.
1147 <title>MODIFY</title>
1150 <command>modify</command>
1152 <arg choice="opt">-H <replaceable>HYPERVISOR_PARAMETERS</replaceable></arg>
1154 <arg choice="opt">-B <replaceable>BACKEND_PARAMETERS</replaceable></arg>
1157 <arg>--net add<replaceable><optional>:options</optional></replaceable></arg>
1158 <arg>--net remove</arg>
1159 <arg>--net <replaceable>N:options</replaceable></arg>
1163 <arg>--disk add:size=<replaceable>SIZE</replaceable></arg>
1164 <arg>--disk remove</arg>
1165 <arg>--disk <replaceable>N</replaceable>:mode=<replaceable>MODE</replaceable></arg>
1171 <arg choice="req"><replaceable>instance</replaceable></arg>
1175 Modifies the memory size, number of vcpus, ip address, MAC
1176 address and/or bridge for an instance. It can also add and
1177 remove disks and NICs to/from the instance. Note that you
1178 need to give at least one of the arguments, otherwise the
1183 The <option>-H</option> option specifies hypervisor options
1184 in the form of <userinput>name=value[,...]</userinput>. For details which options can be specified, see the <command>add</command> command.
1189 add:size=<replaceable>SIZE</replaceable></option> option
1190 adds a disk to the instance. The <option>--disk
1191 remove</option> will remove the last disk of the
1192 instance. The <option>--disk
1193 <replaceable>N</replaceable>:mode=<replaceable>MODE</replaceable></option>
1194 option will change the mode of the Nth disk of the instance
1195 between read-only (<literal>ro</literal>) and read-write
1196 (<literal>rw</literal>).
1201 add:<replaceable>options</replaceable></option> option will
1202 add a new NIC to the instance. The available options are the
1203 same as in the <command>add</command> command (mac, ip,
1204 bridge). The <option>--nice remove</option> will remove the
1205 last NIC of the instance, while the <option>--nic
1206 <replaceable>N</replaceable>:<replaceable>options</replaceable></option>
1207 option will change the parameters of the Nth instance NIC.
1211 The <option>--submit</option> option is used to send the job to
1212 the master daemon but not wait for its completion. The job
1213 ID will be shown so that it can be examined via
1214 <command>gnt-job info</command>.
1218 All the changes take effect at the next restart. If the
1219 instance is running, there is no effect on the instance.
1224 <title>REINSTALL</title>
1227 <command>reinstall</command>
1228 <arg choice="opt">-o <replaceable>os-type</replaceable></arg>
1229 <arg choice="opt">-f <replaceable>force</replaceable></arg>
1230 <arg>--select-os</arg>
1232 <arg choice="req"><replaceable>instance</replaceable></arg>
1236 Reinstalls the operating system on the given instance. The
1237 instance must be stopped when running this command. If the
1238 <option>--os-type</option> is specified, the operating
1243 Since reinstall is potentially dangerous command, the user
1244 will be required to confirm this action, unless the
1245 <option>-f</option> flag is passed.
1249 The <option>--select-os</option> option switches to an
1250 interactive OS reinstall. The user is prompted to select the OS
1251 template from the list of available OS templates.
1255 The <option>--submit</option> option is used to send the job to
1256 the master daemon but not wait for its completion. The job
1257 ID will be shown so that it can be examined via
1258 <command>gnt-job info</command>.
1265 <title>RENAME</title>
1268 <command>rename</command>
1269 <arg>--no-ip-check</arg>
1271 <arg choice="req"><replaceable>instance</replaceable></arg>
1272 <arg choice="req"><replaceable>new_name</replaceable></arg>
1276 Renames the given instance. The instance must be stopped
1277 when running this command. The requirements for the new name
1278 are the same as for adding an instance: the new name must be
1279 resolvable and the IP it resolves to must not be reachable
1280 (in order to prevent duplicate IPs the next time the
1281 instance is started). The IP test can be skipped if the
1282 <option>--no-ip-check</option> option is passed.
1286 The <option>--submit</option> option is used to send the job to
1287 the master daemon but not wait for its completion. The job
1288 ID will be shown so that it can be examined via
1289 <command>gnt-job info</command>.
1297 <title>Starting/stopping/connecting to console</title>
1300 <title>STARTUP</title>
1303 <command>startup</command>
1307 <arg>--force-multiple</arg>
1309 <group choice="opt">
1310 <arg>--instance</arg>
1312 <arg>--primary</arg>
1313 <arg>--secondary</arg>
1317 <arg>-H <option>key=value...</option></arg>
1318 <arg>-B <option>key=value...</option></arg>
1323 rep="repeat"><replaceable>name</replaceable></arg>
1327 Starts one or more instances, depending on the following
1328 options. The four available modes are:
1331 <term><option>--instance</option></term>
1333 <simpara>will start the instances given as arguments
1334 (at least one argument required); this is the default
1341 <simpara>will start the instances who have the given
1342 node as either primary or secondary</simpara>
1346 <term><option>--primary</option></term>
1348 <simpara>will start all instances whose primary node
1349 is in the list of nodes passed as arguments (at least
1350 one node required)</simpara>
1354 <term><option>--secondary</option></term>
1356 <simpara>will start all instances whose secondary node
1357 is in the list of nodes passed as arguments (at least
1358 one node required)</simpara>
1364 <simpara>will start all instances in the cluster (no
1365 arguments accepted)</simpara>
1372 Note that although you can pass more than one selection
1373 option, the last one wins, so in order to guarantee the
1374 desired result, don't pass more than one such option.
1378 Use <option>--force</option> to start even if secondary disks are
1383 The <option>--force-multiple</option> will skip the
1384 interactive confirmation in the case the more than one
1385 instance will be affected.
1389 The <option>-H</option> and <option>-B</option> options
1390 specify extra, temporary hypervisor and backend parameters
1391 that can be used to start an instance with modified
1392 parameters. They can be useful for quick testing without
1393 having to modify an instance back and forth, e.g.:
1395 # gnt-instance start -H root_args="single" instance1
1396 # gnt-instance start -B memory=2048 instance2
1398 The first form will start the instance
1399 <userinput>instance1</userinput> in single-user mode, and
1400 the instance <userinput>instance2</userinput> with 2GB of
1401 RAM (this time only, unless that is the actual instance
1402 memory size already).
1406 The <option>--submit</option> option is used to send the job to
1407 the master daemon but not wait for its completion. The job
1408 ID will be shown so that it can be examined via
1409 <command>gnt-job info</command>.
1415 # gnt-instance start instance1.example.com
1416 # gnt-instance start --node node1.example.com node2.example.com
1417 # gnt-instance start --all
1423 <title>SHUTDOWN</title>
1426 <command>shutdown</command>
1428 <arg>--force-multiple</arg>
1430 <group choice="opt">
1431 <arg>--instance</arg>
1433 <arg>--primary</arg>
1434 <arg>--secondary</arg>
1441 rep="repeat"><replaceable>name</replaceable></arg>
1445 Stops one or more instances. If the instance cannot be
1446 cleanly stopped during a hardcoded interval (currently 2
1447 minutes), it will forcibly stop the instance (equivalent to
1448 switching off the power on a physical machine).
1452 The <option>--instance</option>, <option>--node</option>,
1453 <option>--primary</option>, <option>--secondary</option> and
1454 <option>--all</option> options are similar as for the
1455 <command>startup</command> command and they influence the
1456 actual instances being shutdown.
1460 The <option>--submit</option> option is used to send the job to
1461 the master daemon but not wait for its completion. The job
1462 ID will be shown so that it can be examined via
1463 <command>gnt-job info</command>.
1470 # gnt-instance shutdown instance1.example.com
1471 # gnt-instance shutdown --all
1477 <title>REBOOT</title>
1480 <command>reboot</command>
1482 <arg>--type=<replaceable>REBOOT-TYPE</replaceable></arg>
1484 <arg>--ignore-secondaries</arg>
1486 <arg>--force-multiple</arg>
1488 <group choice="opt">
1489 <arg>--instance</arg>
1491 <arg>--primary</arg>
1492 <arg>--secondary</arg>
1499 rep="repeat"><replaceable>name</replaceable></arg>
1503 Reboots one or more instances. The type of reboot depends on
1504 the value of <option>--type</option>. A soft reboot does a
1505 hypervisor reboot, a hard reboot does a instance stop,
1506 recreates the hypervisor config for the instance and
1507 starts the instance. A full reboot does the equivalent
1508 of <command>gnt-instance shutdown && gnt-instance
1509 startup</command>. The default is hard reboot.
1513 For the hard reboot the option
1514 <option>--ignore-secondaries</option> ignores errors for the
1515 secondary node while re-assembling the instance disks.
1519 The <option>--instance</option>, <option>--node</option>,
1520 <option>--primary</option>, <option>--secondary</option> and
1521 <option>--all</option> options are similar as for the
1522 <command>startup</command> command and they influence the
1523 actual instances being rebooted.
1527 The <option>--force-multiple</option> will skip the
1528 interactive confirmation in the case the more than one
1529 instance will be affected.
1535 # gnt-instance reboot instance1.example.com
1536 # gnt-instance reboot --type=full instance1.example.com
1542 <title>CONSOLE</title>
1544 <command>console</command>
1545 <arg choice="opt">--show-cmd</arg>
1546 <arg choice="req"><replaceable>instance</replaceable></arg>
1550 Connects to the console of the given instance. If the
1551 instance is not up, an error is returned. Use the
1552 <option>--show-cmd</option> option to display the command
1553 instead of executing it.
1557 For HVM instances, this will attempt to connect to the
1558 serial console of the instance. To connect to the
1559 virtualized "physical" console of a HVM instance, use a VNC
1560 client with the connection info from the
1561 <command>info</command> command.
1567 # gnt-instance console instance1.example.com
1575 <title>Disk management</title>
1578 <title>REPLACE-DISKS</title>
1581 <command>replace-disks</command>
1583 <arg choice="req">-p</arg>
1584 <arg>--disks <replaceable>idx</replaceable></arg>
1585 <arg choice="req"><replaceable>instance</replaceable></arg>
1589 <command>replace-disks</command>
1591 <arg choice="req">-s</arg>
1592 <arg>--disks <replaceable>idx</replaceable></arg>
1593 <arg choice="req"><replaceable>instance</replaceable></arg>
1597 <command>replace-disks</command>
1599 <group choice="req">
1600 <arg>--iallocator <replaceable>name</replaceable></arg>
1601 <arg>--new-secondary <replaceable>NODE</replaceable></arg>
1604 <arg choice="req"><replaceable>instance</replaceable></arg>
1608 This command is a generalized form for replacing disks. It
1609 is currently only valid for the mirrored (DRBD) disk
1614 The first form (when passing the <option>-p</option> option)
1615 will replace the disks on the primary, while the second form
1616 (when passing the <option>-s</option> option will replace
1617 the disks on the secondary node. For these two cases (as the
1618 node doesn't change), it is possible to only run the replace
1619 for a subset of the disks, using the option
1620 <option>--disks</option> which takes a list of
1621 comma-delimited disk indices (zero-based),
1622 e.g. <userinput>0,2</userinput> to replace only the first
1627 The third form (when passing either the
1628 <option>--iallocator</option> or the
1629 <option>--new-secondary</option> option) is designed to
1630 change secondary node of the instance. Specifying
1631 <option>--iallocator</option> makes the new secondary be
1632 selected automatically by the specified allocator plugin,
1633 otherwise the new secondary node will be the one chosen
1634 manually via the <option>--new-secondary</option> option.
1638 The <option>--submit</option> option is used to send the job to
1639 the master daemon but not wait for its completion. The job
1640 ID will be shown so that it can be examined via
1641 <command>gnt-job info</command>.
1645 Note that it is not possible to select an offline or drained
1646 node as a new secondary.
1652 <title>ACTIVATE-DISKS</title>
1655 <command>activate-disks</command>
1657 <arg choice="req"><replaceable>instance</replaceable></arg>
1660 Activates the block devices of the given instance. If
1661 successful, the command will show the location and name of
1664 node1.example.com:disk/0:/dev/drbd0
1665 node1.example.com:disk/1:/dev/drbd1
1668 In this example, <emphasis>node1.example.com</emphasis> is
1669 the name of the node on which the devices have been
1670 activated. The <emphasis>disk/0</emphasis> and
1671 <emphasis>disk/1</emphasis> are the Ganeti-names of the
1672 instance disks; how they are visible inside the instance is
1673 hypervisor-specific. <emphasis>/dev/drbd0</emphasis> and
1674 <emphasis>/dev/drbd1</emphasis> are the actual block devices
1675 as visible on the node.
1679 The <option>--submit</option> option is used to send the job to
1680 the master daemon but not wait for its completion. The job
1681 ID will be shown so that it can be examined via
1682 <command>gnt-job info</command>.
1686 Note that it is safe to run this command while the instance
1692 <title>DEACTIVATE-DISKS</title>
1695 <command>deactivate-disks</command>
1697 <arg choice="req"><replaceable>instance</replaceable></arg>
1700 De-activates the block devices of the given instance. Note
1701 that if you run this command for an instance with a drbd
1702 disk template, while it is running, it will not be able to
1703 shutdown the block devices on the primary node, but it will
1704 shutdown the block devices on the secondary nodes, thus
1705 breaking the replication.
1709 The <option>--submit</option> option is used to send the job to
1710 the master daemon but not wait for its completion. The job
1711 ID will be shown so that it can be examined via
1712 <command>gnt-job info</command>.
1718 <title>GROW-DISK</title>
1720 <command>grow-disk</command>
1721 <arg>--no-wait-for-sync</arg>
1723 <arg choice="req"><replaceable>instance</replaceable></arg>
1724 <arg choice="req"><replaceable>disk</replaceable></arg>
1725 <arg choice="req"><replaceable>amount</replaceable></arg>
1729 Grows an instance's disk. This is only possible for
1730 instances having a <literal>plain</literal> or
1731 <literal>drbd</literal> disk template.
1735 Note that this command only change the block device size; it
1736 will not grow the actual filesystems, partitions, etc. that
1737 live on that disk. Usually, you will need to:
1740 <simpara>use <command>gnt-instance grow-disk</command></simpara>
1743 <simpara>reboot the instance (later, at a convenient
1747 <simpara>use a filesystem resizer, such as
1748 <citerefentry> <refentrytitle>ext2online</refentrytitle>
1749 <manvolnum>8</manvolnum> </citerefentry> or
1750 <citerefentry> <refentrytitle>xfs_growfs</refentrytitle>
1751 <manvolnum>8</manvolnum> </citerefentry> to resize the
1752 filesystem, or use <citerefentry>
1753 <refentrytitle>fdisk</refentrytitle>
1754 <manvolnum>8</manvolnum> </citerefentry> to change the
1755 partition table on the disk
1763 The <replaceable>disk</replaceable> argument is the index of
1764 the instance disk to grow. The
1765 <replaceable>amount</replaceable> argument is given either
1766 as a number (and it represents the amount to increase the
1767 disk with in mebibytes) or can be given similar to the
1768 arguments in the create instance operation, with a suffix
1773 Note that the disk grow operation might complete on one node
1774 but fail on the other; this will leave the instance with
1775 different-sized LVs on the two nodes, but this will not
1776 create problems (except for unused space).
1780 If you do not want gnt-instance to wait for the new disk
1781 region to be synced, use the
1782 <option>--no-wait-for-sync</option> option.
1786 The <option>--submit</option> option is used to send the job to
1787 the master daemon but not wait for its completion. The job
1788 ID will be shown so that it can be examined via
1789 <command>gnt-job info</command>.
1793 <para>Example (increase the first disk for instance1 by 16GiB):
1795 # gnt-instance grow-disk instance1.example.com 0 16g
1800 Also note that disk shrinking is not supported; use
1801 <command>gnt-backup export</command> and then
1802 <command>gnt-backup import</command> to reduce the disk size
1810 <title>Recovery</title>
1813 <title>FAILOVER</title>
1816 <command>failover</command>
1818 <arg>--ignore-consistency</arg>
1820 <arg choice="req"><replaceable>instance</replaceable></arg>
1824 Failover will fail the instance over its secondary
1825 node. This works only for instances having a drbd disk
1830 Normally the failover will check the consistency of the
1831 disks before failing over the instance. If you are trying to
1832 migrate instances off a dead node, this will fail. Use the
1833 <option>--ignore-consistency</option> option for this
1834 purpose. Note that this option can be dangerous as errors in
1835 shutting down the instance will be ignored, resulting in
1836 possibly having the instance running on two machines in
1837 parallel (on disconnected DRBD drives).
1841 The <option>--submit</option> option is used to send the job to
1842 the master daemon but not wait for its completion. The job
1843 ID will be shown so that it can be examined via
1844 <command>gnt-job info</command>.
1850 # gnt-instance failover instance1.example.com
1856 <title>MIGRATE</title>
1859 <command>migrate</command>
1861 <arg choice="req">--cleanup</arg>
1862 <arg choice="req"><replaceable>instance</replaceable></arg>
1866 <command>migrate</command>
1868 <arg>--non-live</arg>
1869 <arg choice="req"><replaceable>instance</replaceable></arg>
1873 Migrate will move the instance to its secondary node without
1874 shutdown. It only works for instances having the drbd8 disk
1879 The migration command needs a perfectly healthy instance, as
1880 we rely on the dual-master capability of drbd8 and the disks
1881 of the instance are not allowed to be degraded.
1885 The <option>--non-live</option> option will switch (for the
1886 hypervisors that support it) between a "fully live"
1887 (i.e. the interruption is as minimal as possible) migration
1888 and one in which the instance is frozen, its state saved and
1889 transported to the remote node, and then resumed there. This
1890 all depends on the hypervisor support for two different
1891 methods. In any case, it is not an error to pass this
1892 parameter (it will just be ignored if the hypervisor doesn't
1897 If the <option>--cleanup</option> option is passed, the
1898 operation changes from migration to attempting recovery from
1899 a failed previous migration. In this mode, ganeti checks if
1900 the instance runs on the correct node (and updates its
1901 configuration if not) and ensures the instances's disks are
1902 configured correctly. In this mode, the
1903 <option>--non-live</option> option is ignored.
1907 The option <option>-f</option> will skip the prompting for
1911 Example (and expected output):
1913 # gnt-instance migrate instance1
1914 Migrate will happen to the instance instance1. Note that migration is
1915 **experimental** in this version. This might impact the instance if
1916 anything goes wrong. Continue?
1918 * checking disk consistency between source and target
1919 * ensuring the target is in secondary mode
1920 * changing disks into dual-master mode
1921 - INFO: Waiting for instance instance1 to sync disks.
1922 - INFO: Instance instance1's disks are in sync.
1923 * migrating instance to node2.example.com
1924 * changing the instance's disks on source node to secondary
1925 - INFO: Waiting for instance instance1 to sync disks.
1926 - INFO: Instance instance1's disks are in sync.
1927 * changing the instance's disks to single-master
1939 <title>ADD-TAGS</title>
1942 <command>add-tags</command>
1943 <arg choice="opt">--from <replaceable>file</replaceable></arg>
1944 <arg choice="req"><replaceable>instancename</replaceable></arg>
1946 rep="repeat"><replaceable>tag</replaceable></arg>
1950 Add tags to the given instance. If any of the tags contains
1951 invalid characters, the entire operation will abort.
1954 If the <option>--from</option> option is given, the list of
1955 tags will be extended with the contents of that file (each
1956 line becomes a tag). In this case, there is not need to pass
1957 tags on the command line (if you do, both sources will be
1958 used). A file name of - will be interpreted as stdin.
1963 <title>LIST-TAGS</title>
1966 <command>list-tags</command>
1967 <arg choice="req"><replaceable>instancename</replaceable></arg>
1970 <para>List the tags of the given instance.</para>
1974 <title>REMOVE-TAGS</title>
1976 <command>remove-tags</command>
1977 <arg choice="opt">--from <replaceable>file</replaceable></arg>
1978 <arg choice="req"><replaceable>instancename</replaceable></arg>
1980 rep="repeat"><replaceable>tag</replaceable></arg>
1984 Remove tags from the given instance. If any of the tags are
1985 not existing on the node, the entire operation will abort.
1989 If the <option>--from</option> option is given, the list of
1990 tags will be extended with the contents of that file (each
1991 line becomes a tag). In this case, there is not need to pass
1992 tags on the command line (if you do, both sources will be
1993 used). A file name of - will be interpreted as stdin.
2005 <!-- Keep this comment at the end of the file
2010 sgml-minimize-attributes:nil
2011 sgml-always-quote-attributes:t
2014 sgml-parent-document:nil
2015 sgml-default-dtd-file:nil
2016 sgml-exposed-tags:nil
2017 sgml-local-catalogs:nil
2018 sgml-local-ecat-files:nil