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*]
30 | [\--disk-state *diskstate*]
31 | [\--hypervisor-state *hvstate*]
34 Adds the given node to the cluster.
36 This command is used to join a new node to the cluster. You will
37 have to provide the password for root of the node to be able to add
38 the node in the cluster. The command needs to be run on the Ganeti
41 Note that the command is potentially destructive, as it will
42 forcibly join the specified host the cluster, not paying attention
43 to its current status (it could be already in a cluster, etc.)
45 The ``-s (--secondary-ip)`` is used in dual-home clusters and
46 specifies the new node's IP in the secondary network. See the
47 discussion in **gnt-cluster**(8) for more information.
49 In case you're readding a node after hardware failure, you can use
50 the ``--readd`` parameter. In this case, you don't need to pass the
51 secondary IP again, it will reused from the cluster. Also, the
52 drained and offline flags of the node will be cleared before
55 The ``--force-join`` option is to proceed with adding a node even if it already
56 appears to belong to another cluster. This is used during cluster merging, for
59 The ``-g (--node-group)`` option is used to add the new node into a
60 specific node group, specified by UUID or name. If only one node group
61 exists you can skip this option, otherwise it's mandatory.
63 The ``vm_capable``, ``master_capable``, ``ndparams``, ``diskstate`` and
64 ``hvstate`` options are described in **ganeti**(7), and are used to set
65 the properties of the new node.
67 The command performs some operations that change the state of the master
68 and the new node, like copying certificates and starting the node daemon
69 on the new node, or updating ``/etc/hosts`` on the master node. If the
70 command fails at a later stage, it doesn't undo such changes. This
71 should not be a problem, as a successful run of ``gnt-node add`` will
72 bring everything back in sync.
76 # gnt-node add node5.example.com
77 # gnt-node add -s 192.0.2.5 node5.example.com
78 # gnt-node add -g group2 -s 192.0.2.9 node9.group2.example.com
84 **add-tags** [\--from *file*] {*nodename*} {*tag*...}
86 Add tags to the given node. If any of the tags contains invalid
87 characters, the entire operation will abort.
89 If the ``--from`` option is given, the list of tags will be
90 extended with the contents of that file (each line becomes a tag).
91 In this case, there is not need to pass tags on the command line
92 (if you do, both sources will be used). A file name of - will be
98 | **evacuate** [-f] [\--early-release] [\--submit]
99 | [{-I|\--iallocator} *NAME* \| {-n|\--new-secondary} *destination\_node*]
100 | [{-p|\--primary-only} \| {-s|\--secondary-only} ]
103 This command will move instances away from the given node. If
104 ``--primary-only`` is given, only primary instances are evacuated, with
105 ``--secondary-only`` only secondaries. If neither is given, all
106 instances are evacuated. It works only for instances having a drbd disk
109 The new location for the instances can be specified in two ways:
111 - as a single node for all instances, via the ``-n (--new-secondary)``
114 - or via the ``-I (--iallocator)`` option, giving a script name as
115 parameter (or ``.`` to use the default allocator), so each instance
116 will be in turn placed on the (per the script) optimal node
118 The ``--early-release`` changes the code so that the old storage on
119 node being evacuated is removed early (before the resync is
120 completed) and the internal Ganeti locks are also released for both
121 the current secondary and the new secondary, thus allowing more
122 parallelism in the cluster operation. This should be used only when
123 recovering from a disk failure on the current secondary (thus the
124 old storage is already broken) or when the storage on the primary
125 node is known to be fine (thus we won't need the old storage for
128 Note that this command is equivalent to using per-instance commands for
129 each affected instance individually:
131 - ``--primary-only`` is equivalent to ``gnt-instance
132 failover/migration`` for non-DRBD instances, but for DRBD instances
133 it's different, and usually is a slow process (it will change the
134 primary to another node while keeping the secondary, this requiring
135 data copies, whereas failover/migrate will only toggle the
136 primary/secondary roles, a fast process)
137 - ``--secondary-only`` is equivalent to ``gnt-instance replace-disks``
138 in the secondary node change mode (only valid for DRBD instances)
139 - when neither of the above is done a combination of the two cases is run
141 See **ganeti(7)** for a description of ``--submit`` and other common
146 # gnt-node evacuate -I hail node3.example.com
152 **failover** [-f] [\--ignore-consistency] {*node*}
154 This command will fail over all instances having the given node as
155 primary to their secondary nodes. This works only for instances having
156 a drbd disk template.
158 Normally the failover will check the consistency of the disks before
159 failing over the instance. If you are trying to migrate instances off
160 a dead node, this will fail. Use the ``--ignore-consistency`` option
165 # gnt-node failover node1.example.com
173 Show detailed information about the nodes in the cluster. If you
174 don't give any arguments, all nodes will be shows, otherwise the
175 output will be restricted to the given names.
181 | [\--no-headers] [\--separator=*SEPARATOR*]
182 | [\--units=*UNITS*] [-v] [{-o|\--output} *[+]FIELD,...*]
186 Lists the nodes in the cluster.
188 The ``--no-headers`` option will skip the initial header line. The
189 ``--separator`` option takes an argument which denotes what will be
190 used between the output fields. Both these options are to help
193 The units used to display the numeric values in the output varies,
194 depending on the options given. By default, the values will be
195 formatted in the most appropriate unit. If the ``--separator``
196 option is given, then the values are shown in mebibytes to allow
197 parsing by scripts. In both cases, the ``--units`` option can be
198 used to enforce a given output unit.
200 Queries of nodes will be done in parallel with any running jobs. This might
201 give inconsistent results for the free disk/memory.
203 The ``-v`` option activates verbose mode, which changes the display of
204 special field states (see **ganeti(7)**).
206 The ``-o (--output)`` option takes a comma-separated list of output
207 fields. The available fields and their meaning are:
211 If the value of the option starts with the character ``+``, the new
212 fields will be added to the default list. This allows one to quickly
213 see the default list plus a few other fields, instead of retyping
214 the entire list of fields.
216 Note that some of these fields are known from the configuration of the
217 cluster (e.g. ``name``, ``pinst``, ``sinst``, ``pip``, ``sip``) and thus
218 the master does not need to contact the node for this data (making the
219 listing fast if only fields from this set are selected), whereas the
220 other fields are "live" fields and require a query to the cluster nodes.
222 Depending on the virtualization type and implementation details, the
223 ``mtotal``, ``mnode`` and ``mfree`` fields may have slightly varying
224 meanings. For example, some solutions share the node memory with the
225 pool of memory used for instances (KVM), whereas others have separate
226 memory for the node and for the instances (Xen).
228 If exactly one argument is given and it appears to be a query filter
229 (see **ganeti(7)**), the query result is filtered accordingly. For
230 ambiguous cases (e.g. a single field name as a filter) the ``--filter``
231 (``-F``) option forces the argument to be treated as a filter (e.g.
232 ``gnt-node list -F master_candidate``).
234 If no node names are given, then all nodes are queried. Otherwise,
235 only the given nodes will be listed.
241 **list-drbd** [\--no-headers] [\--separator=*SEPARATOR*] node
243 Lists the mapping of DRBD minors for a given node. This outputs a static
244 list of fields (it doesn't accept the ``--output`` option), as follows:
247 The (full) name of the node we are querying
251 The instance the DRBD minor belongs to
253 The disk index that the DRBD minor belongs to
255 Either ``primary`` or ``secondary``, denoting the role of the node for
256 the instance (note: this is not the live status of the DRBD device,
257 but the configuration value)
259 The node that the minor is connected to on the other end
261 This command can be used as a reverse lookup (from node and minor) to a
262 given instance, which can be useful when debugging DRBD issues.
264 Note that this command queries Ganeti via :manpage:`ganeti-confd(8)`, so
265 it won't be available if support for ``confd`` has not been enabled at
266 build time; furthermore, in Ganeti 2.6 this is only available via the
267 Haskell version of confd (again selected at build time).
272 **list-fields** [field...]
274 Lists available fields for nodes.
280 **list-tags** {*nodename*}
282 List the tags of the given node.
287 | **migrate** [-f] [\--non-live] [\--migration-mode=live\|non-live]
288 | [\--ignore-ipolicy] [\--submit] {*node*}
290 This command will migrate all instances having the given node as
291 primary to their secondary nodes. This works only for instances
292 having a drbd disk template.
294 As for the **gnt-instance migrate** command, the options
295 ``--no-live``, ``--migration-mode`` and ``--no-runtime-changes``
296 can be given to influence the migration type.
298 If ``--ignore-ipolicy`` is given any instance policy violations
299 occurring during this operation are ignored.
301 See **ganeti(7)** for a description of ``--submit`` and other common
306 # gnt-node migrate node1.example.com
312 | **modify** [-f] [\--submit]
313 | [{-C|\--master-candidate} ``yes|no``]
314 | [{-D|\--drained} ``yes|no``] [{-O|\--offline} ``yes|no``]
315 | [\--master-capable=``yes|no``] [\--vm-capable=``yes|no``] [\--auto-promote]
316 | [{-s|\--secondary-ip} *secondary_ip*]
317 | [\--node-parameters *ndparams*]
318 | [\--node-powered=``yes|no``]
319 | [\--hypervisor-state *hvstate*]
320 | [\--disk-state *diskstate*]
323 This command changes the role of the node. Each options takes
324 either a literal yes or no, and only one option should be given as
325 yes. The meaning of the roles and flags are described in the
326 manpage **ganeti(7)**.
328 The option ``--node-powered`` can be used to modify state-of-record if
329 it doesn't reflect the reality anymore.
331 In case a node is demoted from the master candidate role, the
332 operation will be refused unless you pass the ``--auto-promote``
333 option. This option will cause the operation to lock all cluster nodes
334 (thus it will not be able to run in parallel with most other jobs),
335 but it allows automated maintenance of the cluster candidate pool. If
336 locking all cluster node is too expensive, another option is to
337 promote manually another node to master candidate before demoting the
340 Example (setting a node offline, which will demote it from master
341 candidate role if is in that role)::
343 # gnt-node modify --offline=yes node1.example.com
345 The ``-s (--secondary-ip)`` option can be used to change the node's
346 secondary ip. No drbd instances can be running on the node, while this
347 operation is taking place. Remember that the secondary ip must be
348 reachable from the master secondary ip, when being changed, so be sure
349 that the node has the new IP already configured and active. In order to
350 convert a cluster from single homed to multi-homed or vice versa
351 ``--force`` is needed as well, and the target node for the first change
354 See **ganeti(7)** for a description of ``--submit`` and other common
357 Example (setting the node back to online and master candidate)::
359 # gnt-node modify --offline=no --master-candidate=yes node1.example.com
365 **remove** {*nodename*}
367 Removes a node from the cluster. Instances must be removed or
368 migrated to another cluster before.
372 # gnt-node remove node5.example.com
378 **remove-tags** [\--from *file*] {*nodename*} {*tag*...}
380 Remove tags from the given node. If any of the tags are not
381 existing on the node, the entire operation will abort.
383 If the ``--from`` option is given, the list of tags to be removed will
384 be extended with the contents of that file (each line becomes a tag).
385 In this case, there is not need to pass tags on the command line (if
386 you do, tags from both sources will be removed). A file name of - will
387 be interpreted as stdin.
392 | **volumes** [\--no-headers] [\--human-readable]
393 | [\--separator=*SEPARATOR*] [{-o|\--output} *FIELDS*]
396 Lists all logical volumes and their physical disks from the node(s)
399 The ``--no-headers`` option will skip the initial header line. The
400 ``--separator`` option takes an argument which denotes what will be
401 used between the output fields. Both these options are to help
404 The units used to display the numeric values in the output varies,
405 depending on the options given. By default, the values will be
406 formatted in the most appropriate unit. If the ``--separator``
407 option is given, then the values are shown in mebibytes to allow
408 parsing by scripts. In both cases, the ``--units`` option can be
409 used to enforce a given output unit.
411 The ``-o (--output)`` option takes a comma-separated list of output
412 fields. The available fields and their meaning are:
415 the node name on which the volume exists
418 the physical drive (on which the LVM physical volume lives)
421 the volume group name
424 the logical volume name
427 the logical volume size
430 The name of the instance to which this volume belongs, or (in case
431 it's an orphan volume) the character "-"
436 # gnt-node volumes node5.example.com
437 Node PhysDev VG Name Size Instance
438 node1.example.com /dev/hdc1 xenvg instance1.example.com-sda_11000.meta 128 instance1.example.com
439 node1.example.com /dev/hdc1 xenvg instance1.example.com-sda_11001.data 256 instance1.example.com
445 | **list-storage** [\--no-headers] [\--human-readable]
446 | [\--separator=*SEPARATOR*] [\--storage-type=*STORAGE\_TYPE*]
447 | [{-o|\--output} *FIELDS*]
450 Lists the available storage units and their details for the given
453 The ``--no-headers`` option will skip the initial header line. The
454 ``--separator`` option takes an argument which denotes what will be
455 used between the output fields. Both these options are to help
458 The units used to display the numeric values in the output varies,
459 depending on the options given. By default, the values will be
460 formatted in the most appropriate unit. If the ``--separator``
461 option is given, then the values are shown in mebibytes to allow
462 parsing by scripts. In both cases, the ``--units`` option can be
463 used to enforce a given output unit.
465 The ``--storage-type`` option can be used to choose a storage unit
466 type. Possible choices are lvm-pv, lvm-vg or file.
468 The ``-o (--output)`` option takes a comma-separated list of output
469 fields. The available fields and their meaning are:
472 the node name on which the volume exists
475 the type of the storage unit (currently just what is passed in via
479 the path/identifier of the storage unit
482 total size of the unit; for the file type see a note below
485 used space in the unit; for the file type see a note below
491 whether we the unit is available for allocation (only lvm-pv can
492 change this setting, the other types always report true)
495 Note that for the "file" type, the total disk space might not equal
496 to the sum of used and free, due to the method Ganeti uses to
497 compute each of them. The total and free values are computed as the
498 total and free space values for the filesystem to which the
499 directory belongs, but the used space is computed from the used
500 space under that directory *only*, which might not be necessarily
501 the root of the filesystem, and as such there could be files
502 outside the file storage directory using disk space and causing a
503 mismatch in the values.
507 node1# gnt-node list-storage node2
508 Node Type Name Size Used Free Allocatable
509 node2 lvm-pv /dev/sda7 673.8G 1.5G 672.3G Y
510 node2 lvm-pv /dev/sdb1 698.6G 0M 698.6G Y
516 | **modify-storage** [\--allocatable={yes|no}] [\--submit]
517 | {*node*} {*storage-type*} {*volume-name*}
519 Modifies storage volumes on a node. Only LVM physical volumes can
520 be modified at the moment. They have a storage type of "lvm-pv".
524 # gnt-node modify-storage --allocatable no node5.example.com lvm-pv /dev/sdb1
530 | **repair-storage** [\--ignore-consistency] ]\--submit]
531 | {*node*} {*storage-type*} {*volume-name*}
533 Repairs a storage volume on a node. Only LVM volume groups can be
534 repaired at this time. They have the storage type "lvm-vg".
536 On LVM volume groups, **repair-storage** runs ``vgreduce
541 **Caution:** Running this command can lead to data loss. Use it with
544 The ``--ignore-consistency`` option will ignore any inconsistent
545 disks (on the nodes paired with this one). Use of this option is
546 most likely to lead to data-loss.
550 # gnt-node repair-storage node5.example.com lvm-vg xenvg
556 **powercycle** [\--yes] [\--force] [\--submit] {*node*}
558 This command (tries to) forcefully reboot a node. It is a command
559 that can be used if the node environment is broken, such that the
560 admin can no longer login over SSH, but the Ganeti node daemon is
563 Note that this command is not guaranteed to work; it depends on the
564 hypervisor how effective is the reboot attempt. For Linux, this
565 command requires the kernel option ``CONFIG_MAGIC_SYSRQ`` to be
568 The ``--yes`` option can be used to skip confirmation, while the
569 ``--force`` option is needed if the target node is the master
572 See **ganeti(7)** for a description of ``--submit`` and other common
578 **power** [``--force``] [``--ignore-status``] [``--all``]
579 [``--power-delay``] on|off|cycle|status [*nodes*]
581 This command calls out to out-of-band management to change the power
582 state of given node. With ``status`` you get the power status as reported
583 by the out-of-band management script.
585 Note that this command will only work if the out-of-band functionality
586 is configured and enabled on the cluster. If this is not the case,
587 please use the **powercycle** command above.
589 Using ``--force`` you skip the confirmation to do the operation.
590 Currently this only has effect on ``off`` and ``cycle``. On those two
591 you can *not* operate on the master. However, the command will provide
592 you with the command to invoke to operate on the master nerver-mind.
593 This is considered harmful and Ganeti does not support the use of it.
595 Providing ``--ignore-status`` will ignore the offline=N state of a node
596 and continue with power off.
598 ``--power-delay`` specifies the time in seconds (factions allowed)
599 waited between powering on the next node. This is by default 2 seconds
600 but can increased if needed with this option.
602 *nodes* are optional. If not provided it will call out for every node in
603 the cluster. Except for the ``off`` and ``cycle`` command where you've
604 to explicit use ``--all`` to select all.
612 This command calls out to out-of-band management to ask for the health status
613 of all or given nodes. The health contains the node name and then the items
614 element with their status in a ``item=status`` manner. Where ``item`` is script
615 specific and ``status`` can be one of ``OK``, ``WARNING``, ``CRITICAL`` or
616 ``UNKNOWN``. Items with status ``WARNING`` or ``CRITICAL`` are logged and
617 annotated in the command line output.
619 .. vim: set textwidth=72 :