1 gnt-node(8) Ganeti | Version @GANETI_VERSION@
2 =============================================
7 gnt-node - Node administration
12 **gnt-node** {command} [arguments...]
17 The **gnt-node** is used for managing the (physical) nodes in the
26 | **add** [--readd] [{-s|--secondary-ip} *secondary\_ip*]
27 | [{-g|--node-group} *nodegroup*]
28 | [--master-capable=``yes|no``] [--vm-capable=``yes|no``]
29 | [--node-parameters *ndparams*]
32 Adds the given node to the cluster.
34 This command is used to join a new node to the cluster. You will
35 have to provide the password for root of the node to be able to add
36 the node in the cluster. The command needs to be run on the Ganeti
39 Note that the command is potentially destructive, as it will
40 forcibly join the specified host the cluster, not paying attention
41 to its current status (it could be already in a cluster, etc.)
43 The ``-s (--secondary-ip)`` is used in dual-home clusters and
44 specifies the new node's IP in the secondary network. See the
45 discussion in **gnt-cluster**(8) for more information.
47 In case you're readding a node after hardware failure, you can use
48 the ``--readd`` parameter. In this case, you don't need to pass the
49 secondary IP again, it will reused from the cluster. Also, the
50 drained and offline flags of the node will be cleared before
53 The ``--force-join`` option is to proceed with adding a node even if it already
54 appears to belong to another cluster. This is used during cluster merging, for
57 The ``-g (--node-group)`` option is used to add the new node into a
58 specific node group, specified by UUID or name. If only one node group
59 exists you can skip this option, otherwise it's mandatory.
61 The ``vm_capable``, ``master_capable`` and ``ndparams`` options are
62 described in **ganeti**(7), and are used to set the properties of the
67 # gnt-node add node5.example.com
68 # gnt-node add -s 192.0.2.5 node5.example.com
69 # gnt-node add -g group2 -s 192.0.2.9 node9.group2.example.com
75 **add-tags** [--from *file*] {*nodename*} {*tag*...}
77 Add tags to the given node. If any of the tags contains invalid
78 characters, the entire operation will abort.
80 If the ``--from`` option is given, the list of tags will be
81 extended with the contents of that file (each line becomes a tag).
82 In this case, there is not need to pass tags on the command line
83 (if you do, both sources will be used). A file name of - will be
89 **evacuate** [-f] [--early-release] [--iallocator *NAME* \|
90 --new-secondary *destination\_node*]
91 [--primary-only \| --secondary-only] [--early-release] {*node*}
93 This command will move instances away from the given node. If
94 ``--primary-only`` is given, only primary instances are evacuated, with
95 ``--secondary-only`` only secondaries. If neither is given, all
96 instances are evacuated. It works only for instances having a drbd disk
99 The new location for the instances can be specified in two ways:
101 - as a single node for all instances, via the ``-n (--new-secondary)``
104 - or via the ``-I (--iallocator)`` option, giving a script name as
105 parameter, so each instance will be in turn placed on the (per the
108 The ``--early-release`` changes the code so that the old storage on
109 node being evacuated is removed early (before the resync is
110 completed) and the internal Ganeti locks are also released for both
111 the current secondary and the new secondary, thus allowing more
112 parallelism in the cluster operation. This should be used only when
113 recovering from a disk failure on the current secondary (thus the
114 old storage is already broken) or when the storage on the primary
115 node is known to be fine (thus we won't need the old storage for
120 # gnt-node evacuate -I hail node3.example.com
126 **failover** [-f] [--ignore-consistency] {*node*}
128 This command will fail over all instances having the given node as
129 primary to their secondary nodes. This works only for instances having
130 a drbd disk template.
132 Normally the failover will check the consistency of the disks before
133 failing over the instance. If you are trying to migrate instances off
134 a dead node, this will fail. Use the ``--ignore-consistency`` option
139 # gnt-node failover node1.example.com
147 Show detailed information about the nodes in the cluster. If you
148 don't give any arguments, all nodes will be shows, otherwise the
149 output will be restricted to the given names.
155 | [--no-headers] [--separator=*SEPARATOR*]
156 | [--units=*UNITS*] [-v] [{-o|--output} *[+]FIELD,...*]
160 Lists the nodes in the cluster.
162 The ``--no-headers`` option will skip the initial header line. The
163 ``--separator`` option takes an argument which denotes what will be
164 used between the output fields. Both these options are to help
167 The units used to display the numeric values in the output varies,
168 depending on the options given. By default, the values will be
169 formatted in the most appropriate unit. If the ``--separator``
170 option is given, then the values are shown in mebibytes to allow
171 parsing by scripts. In both cases, the ``--units`` option can be
172 used to enforce a given output unit.
174 Queries of nodes will be done in parallel with any running jobs. This might
175 give inconsistent results for the free disk/memory.
177 The ``-v`` option activates verbose mode, which changes the display of
178 special field states (see **ganeti(7)**).
180 The ``-o (--output)`` option takes a comma-separated list of output
181 fields. The available fields and their meaning are:
185 If the value of the option starts with the character ``+``, the new
186 fields will be added to the default list. This allows one to quickly
187 see the default list plus a few other fields, instead of retyping
188 the entire list of fields.
190 Note that some of these fields are known from the configuration of the
191 cluster (e.g. ``name``, ``pinst``, ``sinst``, ``pip``, ``sip``) and thus
192 the master does not need to contact the node for this data (making the
193 listing fast if only fields from this set are selected), whereas the
194 other fields are "live" fields and require a query to the cluster nodes.
196 Depending on the virtualization type and implementation details, the
197 ``mtotal``, ``mnode`` and ``mfree`` fields may have slighly varying
198 meanings. For example, some solutions share the node memory with the
199 pool of memory used for instances (KVM), whereas others have separate
200 memory for the node and for the instances (Xen).
202 If exactly one argument is given and it appears to be a query filter
203 (see **ganeti(7)**), the query result is filtered accordingly. For
204 ambiguous cases (e.g. a single field name as a filter) the ``--filter``
205 (``-F``) option forces the argument to be treated as a filter (e.g.
206 ``gnt-node list -F master_candidate``).
208 If no node names are given, then all nodes are queried. Otherwise,
209 only the given nodes will be listed.
215 **list-fields** [field...]
217 Lists available fields for nodes.
223 **list-tags** {*nodename*}
225 List the tags of the given node.
230 **migrate** [-f] [--non-live] [--migration-mode=live\|non-live]
233 This command will migrate all instances having the given node as
234 primary to their secondary nodes. This works only for instances
235 having a drbd disk template.
237 As for the **gnt-instance migrate** command, the options
238 ``--no-live`` and ``--migration-mode`` can be given to influence
243 # gnt-node migrate node1.example.com
249 | **modify** [-f] [--submit]
250 | [{-C|--master-candidate} ``yes|no``]
251 | [{-D|--drained} ``yes|no``] [{-O|--offline} ``yes|no``]
252 | [--master-capable=``yes|no``] [--vm-capable=``yes|no``] [--auto-promote]
253 | [{-s|--secondary-ip} *secondary_ip*]
254 | [--node-parameters *ndparams*]
255 | [--node-powered=``yes|no``]
258 This command changes the role of the node. Each options takes
259 either a literal yes or no, and only one option should be given as
260 yes. The meaning of the roles and flags are described in the
261 manpage **ganeti**(7).
263 The option ``--node-powered`` can be used to modify state-of-record if
264 it doesn't reflect the reality anymore.
266 In case a node is demoted from the master candidate role, the
267 operation will be refused unless you pass the ``--auto-promote``
268 option. This option will cause the operation to lock all cluster nodes
269 (thus it will not be able to run in parallel with most other jobs),
270 but it allows automated maintenance of the cluster candidate pool. If
271 locking all cluster node is too expensive, another option is to
272 promote manually another node to master candidate before demoting the
275 Example (setting a node offline, which will demote it from master
276 candidate role if is in that role)::
278 # gnt-node modify --offline=yes node1.example.com
280 The ``-s (--secondary-ip)`` option can be used to change the node's
281 secondary ip. No drbd instances can be running on the node, while this
282 operation is taking place.
284 Example (setting the node back to online and master candidate)::
286 # gnt-node modify --offline=no --master-candidate=yes node1.example.com
292 **remove** {*nodename*}
294 Removes a node from the cluster. Instances must be removed or
295 migrated to another cluster before.
299 # gnt-node remove node5.example.com
305 **remove-tags** [--from *file*] {*nodename*} {*tag*...}
307 Remove tags from the given node. If any of the tags are not
308 existing on the node, the entire operation will abort.
310 If the ``--from`` option is given, the list of tags to be removed will
311 be extended with the contents of that file (each line becomes a tag).
312 In this case, there is not need to pass tags on the command line (if
313 you do, tags from both sources will be removed). A file name of - will
314 be interpreted as stdin.
319 | **volumes** [--no-headers] [--human-readable]
320 | [--separator=*SEPARATOR*] [{-o|--output} *FIELDS*]
323 Lists all logical volumes and their physical disks from the node(s)
326 The ``--no-headers`` option will skip the initial header line. The
327 ``--separator`` option takes an argument which denotes what will be
328 used between the output fields. Both these options are to help
331 The units used to display the numeric values in the output varies,
332 depending on the options given. By default, the values will be
333 formatted in the most appropriate unit. If the ``--separator``
334 option is given, then the values are shown in mebibytes to allow
335 parsing by scripts. In both cases, the ``--units`` option can be
336 used to enforce a given output unit.
338 The ``-o (--output)`` option takes a comma-separated list of output
339 fields. The available fields and their meaning are:
342 the node name on which the volume exists
345 the physical drive (on which the LVM physical volume lives)
348 the volume group name
351 the logical volume name
354 the logical volume size
357 The name of the instance to which this volume belongs, or (in case
358 it's an orphan volume) the character "-"
363 # gnt-node volumes node5.example.com
364 Node PhysDev VG Name Size Instance
365 node1.example.com /dev/hdc1 xenvg instance1.example.com-sda_11000.meta 128 instance1.example.com
366 node1.example.com /dev/hdc1 xenvg instance1.example.com-sda_11001.data 256 instance1.example.com
372 | **list-storage** [--no-headers] [--human-readable]
373 | [--separator=*SEPARATOR*] [--storage-type=*STORAGE\_TYPE*]
374 | [{-o|--output} *FIELDS*]
377 Lists the available storage units and their details for the given
380 The ``--no-headers`` option will skip the initial header line. The
381 ``--separator`` option takes an argument which denotes what will be
382 used between the output fields. Both these options are to help
385 The units used to display the numeric values in the output varies,
386 depending on the options given. By default, the values will be
387 formatted in the most appropriate unit. If the ``--separator``
388 option is given, then the values are shown in mebibytes to allow
389 parsing by scripts. In both cases, the ``--units`` option can be
390 used to enforce a given output unit.
392 The ``--storage-type`` option can be used to choose a storage unit
393 type. Possible choices are lvm-pv, lvm-vg or file.
395 The ``-o (--output)`` option takes a comma-separated list of output
396 fields. The available fields and their meaning are:
399 the node name on which the volume exists
402 the type of the storage unit (currently just what is passed in via
406 the path/identifier of the storage unit
409 total size of the unit; for the file type see a note below
412 used space in the unit; for the file type see a note below
418 whether we the unit is available for allocation (only lvm-pv can
419 change this setting, the other types always report true)
422 Note that for the "file" type, the total disk space might not equal
423 to the sum of used and free, due to the method Ganeti uses to
424 compute each of them. The total and free values are computed as the
425 total and free space values for the filesystem to which the
426 directory belongs, but the used space is computed from the used
427 space under that directory *only*, which might not be necessarily
428 the root of the filesystem, and as such there could be files
429 outside the file storage directory using disk space and causing a
430 mismatch in the values.
434 node1# gnt-node list-storage node2
435 Node Type Name Size Used Free Allocatable
436 node2 lvm-pv /dev/sda7 673.8G 1.5G 672.3G Y
437 node2 lvm-pv /dev/sdb1 698.6G 0M 698.6G Y
443 **modify-storage** [``--allocatable=yes|no``]
444 {*node*} {*storage-type*} {*volume-name*}
446 Modifies storage volumes on a node. Only LVM physical volumes can
447 be modified at the moment. They have a storage type of "lvm-pv".
451 # gnt-node modify-storage --allocatable no node5.example.com lvm-pv /dev/sdb1
457 **repair-storage** [--ignore-consistency] {*node*} {*storage-type*}
460 Repairs a storage volume on a node. Only LVM volume groups can be
461 repaired at this time. They have the storage type "lvm-vg".
463 On LVM volume groups, **repair-storage** runs "vgreduce
468 **Caution:** Running this command can lead to data loss. Use it with
471 The ``--ignore-consistency`` option will ignore any inconsistent
472 disks (on the nodes paired with this one). Use of this option is
473 most likely to lead to data-loss.
477 # gnt-node repair-storage node5.example.com lvm-vg xenvg
483 **powercycle** [``--yes``] [``--force``] {*node*}
485 This commands (tries to) forcefully reboot a node. It is a command
486 that can be used if the node environemnt is broken, such that the
487 admin can no longer login over ssh, but the Ganeti node daemon is
490 Note that this command is not guaranteed to work; it depends on the
491 hypervisor how effective is the reboot attempt. For Linux, this
492 command require that the kernel option CONFIG\_MAGIC\_SYSRQ is
495 The ``--yes`` option can be used to skip confirmation, while the
496 ``--force`` option is needed if the target node is the master
502 **power** [``--force``] [``--ignore-status``] [``--all``]
503 [``--power-delay``] on|off|cycle|status [*nodes*]
505 This commands calls out to out-of-band management to change the power
506 state of given node. With ``status`` you get the power status as reported
507 by the out-of-band managment script.
509 Note that this command will only work if the out-of-band functionality
510 is configured and enabled on the cluster. If this is not the case,
511 please use the **powercycle** command above.
513 Using ``--force`` you skip the confirmation to do the operation.
514 Currently this only has effect on ``off`` and ``cycle``. On those two
515 you can *not* operate on the master. However, the command will provide
516 you with the command to invoke to operate on the master nerver-mind.
517 This is considered harmful and Ganeti does not support the use of it.
519 Providing ``--ignore-status`` will ignore the offline=N state of a node
520 and continue with power off.
522 ``--power-delay`` specifies the time in seconds (factions allowed)
523 waited between powering on the next node. This is by default 2 seconds
524 but can increased if needed with this option.
526 *nodes* are optional. If not provided it will call out for every node in
527 the cluster. Except for the ``off`` and ``cycle`` command where you've
528 to explicit use ``--all`` to select all.
536 This commands calls out to out-pf-band management to ask for the health status
537 of all or given nodes. The health contains the node name and then the items
538 element with their status in a ``item=status`` manner. Where ``item`` is script
539 specific and ``status`` can be one of ``OK``, ``WARNING``, ``CRITICAL`` or
540 ``UNKNOWN``. Items with status ``WARNING`` or ``CRITICAL`` are logged and
541 annotated in the command line output.
543 .. vim: set textwidth=72 :