1VGCHANGE(8) System Manager's Manual VGCHANGE(8)
2
6 vgchange — change attributes of a volume group
7
9 vgchange [--addtag Tag] [--alloc AllocationPolicy] [-A|--autobackup
10 {y|n}] [-a|--activate [a|e|s|l] {y|n}] [--activationmode {com‐
11 plete|degraded|partial}] [-K|--ignoreactivationskip] [--monitor {y|n}]
12 [--poll {y|n}] [-c|--clustered {y|n}] [-u|--uuid] [--commandprofile
13 ProfileName] [-d|--debug] [--deltag Tag] [--detachprofile] [-h|--help]
14 [--ignorelockingfailure] [--ignoremonitoring] [--ignoreskippedcluster]
15 [--sysinit] [--noudevsync] [--lock-start] [--lock-stop] [--lock-type
16 LockType] [-l|--logicalvolume MaxLogicalVolumes] [-p|--maxphysicalvol‐
17 umes MaxPhysicalVolumes] [--metadataprofile ProfileName] [--[vg]meta‐
18 datacopies] NumberOfCopies|unmanaged|all] [-P|--partial] [-s|--physi‐
19 calextentsize PhysicalExtentSize[bBsSkKmMgGtTpPeE]] [-S|--select Selec‐
20 tion] [--systemid SystemID] [--refresh] [-t|--test] [-v|--verbose]
21 [--version] [-x|--resizeable {y|n}] [VolumeGroupName...]
22
24 vgchange allows you to change the attributes of one or more volume
25 groups. Its main purpose is to activate and deactivate VolumeGroupName,
26 or all volume groups if none is specified. Only active volume groups
27 are subject to changes and allow access to their logical volumes. [Not
28 yet implemented: During volume group activation, if vgchange recognizes
29 snapshot logical volumes which were dropped because they ran out of
30 space, it displays a message informing the administrator that such
31 snapshots should be removed (see lvremove(8)). ]
32
34 See lvm(8) for common options.
35 -A, --autobackup {y|n}
37 Controls automatic backup of metadata after the change. See
38 vgcfgbackup(8). Default is yes.
39 -a, --activate [a|e|s|l]{y|n}
41 Controls the availability of the logical volumes in the volume
42 group for input/output. In other words, makes the logical vol‐
43 umes known/unknown to the kernel. If autoactivation option is
44 used (-aay), each logical volume in the volume group is acti‐
45 vated only if it matches an item in the activation/auto_activa‐
46 tion_volume_list set in lvm.conf. If this list is not set, then
47 all volumes are considered for activation. The -aay option
48 should be also used during system boot so it's possible to
49 select which volumes to activate using the activation/auto_acti‐
50 vation_volume_list settting.
51 Activation of a logical volume creates a symbolic link /dev/Vol‐
53 umeGroupName/LogicalVolumeName pointing to the device node.
54 This link is removed on deactivation. All software and scripts
55 should access the device through this symbolic link and present
56 this as the name of the device. The location and name of the
57 underlying device node may depend on the distribution and con‐
58 figuration (e.g. udev) and might change from release to release.
59 In a clustered VG, clvmd is used for activation, and the follow‐
61 ing options are possible:
62 With -aey, clvmd activates the LV in exclusive mode (with an
64 exclusive lock), allowing a single node to activate the LV.
65 With -asy, clvmd activates the LV in shared mode (with a shared
67 lock), allowing multiple nodes to activate the LV concurrently.
68 If the LV type prohibits shared access, such as an LV with a
69 snapshot, the 's' option is ignored and an exclusive lock is
70 used.
71 With -ay (no mode specified), clvmd activates the LV in shared
73 mode if the LV type allows concurrent access, such as a linear
74 LV. Otherwise, clvmd activates the LV in exclusive mode.
75 With -aey, -asy, and -ay, clvmd attempts to activate the LV on
77 all nodes. If exclusive mode is used, then only one of the
78 nodes will be successful.
79 With -an, clvmd attempts to deactivate the LV on all nodes.
81 With -aly, clvmd activates the LV only on the local node, and
83 -aln deactivates only on the local node. If the LV type allows
84 concurrent access, then shared mode is used, otherwise exclu‐
85 sive.
86 LVs with snapshots are always activated exclusively because they
88 can only be used on one node at once.
89 For local VGs, -ay, -aey, and -asy are all equivalent.
91 In a shared VG, lvmlockd is used for locking if LVM is compiled
93 with lockd support, and the following options are possible:
94 With -aey, the command activates the LV in exclusive mode,
96 allowing a single host to activate the LV (the host running the
97 command). Before activating the LV, the command uses lvmlockd
98 to acquire an exclusive lock on the LV. If the lock cannot be
99 acquired, the LV is not activated and an error is reported.
100 This would happen if the LV is active on another host.
101 With -asy, the command activates the LV in shared mode, allowing
103 multiple hosts to activate the LV concurrently. Before activat‐
104 ing the LV, the command uses lvmlockd to acquire a shared lock
105 on the LV. If the lock cannot be acquired, the LV is not acti‐
106 vated and an error is reported. This would happen if the LV is
107 active exclusively on another host. If the LV type prohibits
108 shared access, such as a snapshot, the command will report an
109 error and fail.
110 With -an, the command deactivates the LV on the host running the
112 command. After deactivating the LV, the command uses lvmlockd
113 to release the current lock on the LV.
114 With lvmlockd, an unspecified mode is always exclusive, -ay
116 defaults to -aey.
117 --activationmode {complete|degraded|partial}
120 The activation mode determines whether logical volumes are
121 allowed to activate when there are physical volumes missing
122 (e.g. due to a device failure). complete is the most restric‐
123 tive; allowing only those logical volumes to be activated that
124 are not affected by the missing PVs. degraded allows RAID logi‐
125 cal volumes to be activated even if they have PVs missing.
126 (Note that the "mirror" segment type is not considered a RAID
127 logical volume. The "raid1" segment type should be used
128 instead.) Finally, partial allows any logical volume to be
129 activated even if portions are missing due to a missing or
130 failed PV. This last option should only be used when performing
131 recovery or repair operations. degraded is the default mode.
132 To change it, modify activation_mode in lvm.conf(5).
133 -K, --ignoreactivationskip
135 Ignore the flag to skip Logical Volumes during activation.
136 -c, --clustered {y|n}
138 If clustered locking is enabled, this indicates whether this
139 Volume Group is shared with other nodes in the cluster or
140 whether it contains only local disks that are not visible on the
141 other nodes. If the cluster infrastructure is unavailable on a
142 particular node at a particular time, you may still be able to
143 use Volume Groups that are not marked as clustered.
144 --detachprofile
146 Detach any metadata configuration profiles attached to given
147 Volume Groups. See lvm.conf(5) for more information about meta‐
148 data profiles.
149 -u, --uuid
151 Generate new random UUID for specified Volume Groups.
152 --monitor {y|n}
154 Start or stop monitoring a mirrored or snapshot logical volume
155 with dmeventd, if it is installed. If a device used by a moni‐
156 tored mirror reports an I/O error, the failure is handled
157 according to mirror_image_fault_policy and mirror_log_fault_pol‐
158 icy set in lvm.conf(5).
159 --poll {y|n}
161 Without polling a logical volume's backgrounded transformation
162 process will never complete. If there is an incomplete pvmove
163 or lvconvert (for example, on rebooting after a crash), use
164 --poll y to restart the process from its last checkpoint. How‐
165 ever, it may not be appropriate to immediately poll a logical
166 volume when it is activated, use --poll n to defer and then
167 --poll y to restart the process.
168 --sysinit
170 Indicates that vgchange(8) is being invoked from early system
171 initialisation scripts (e.g. rc.sysinit or an initrd), before
172 writeable filesystems are available. As such, some functionality
173 needs to be disabled and this option acts as a shortcut which
174 selects an appropriate set of options. Currently this is equiva‐
175 lent to using --ignorelockingfailure, --ignoremonitoring, --poll
176 n and setting LVM_SUPPRESS_LOCKING_FAILURE_MESSAGES environment
177 variable.
178 If --sysinit is used in conjunction with lvmetad(8) enabled and
180 running, autoactivation is preferred over manual activation via
181 direct vgchange call. Logical volumes are autoactivated accord‐
182 ing to auto_activation_volume_list set in lvm.conf(5).
183 --noudevsync
185 Disable udev synchronisation. The process will not wait for
186 notification from udev. It will continue irrespective of any
187 possible udev processing in the background. You should only use
188 this if udev is not running or has rules that ignore the devices
189 LVM2 creates.
190 --ignoremonitoring
192 Make no attempt to interact with dmeventd unless --monitor is
193 specified. Do not use this if dmeventd is already monitoring a
194 device.
195 --lock-start
197 Start the lockspace of a shared VG in lvmlockd. lvmlockd locks
198 becomes available for the VG, allowing LVM to use the VG. See
199 lvmlockd(8).
200 --lock-stop
202 Stop the lockspace of a shared VG in lvmlockd. lvmlockd locks
203 become unavailable for the VG, preventing LVM from using the VG.
204 See lvmlockd(8).
205 --lock-type LockType
207 Change the VG lock type to or from a shared lock type used with
208 lvmlockd. See lvmlockd(8).
209 -l, --logicalvolume MaxLogicalVolumes
211 Changes the maximum logical volume number of an existing inac‐
212 tive volume group.
213 -p, --maxphysicalvolumes MaxPhysicalVolumes
215 Changes the maximum number of physical volumes that can belong
216 to this volume group. For volume groups with metadata in lvm1
217 format, the limit is 255. If the metadata uses lvm2 format, the
218 value 0 removes this restriction: there is then no limit. If
219 you have a large number of physical volumes in a volume group
220 with metadata in lvm2 format, for tool performance reasons, you
221 should consider some use of --pvmetadatacopies 0 as described in
222 pvcreate(8), and/or use --vgmetadatacopies.
223 --metadataprofile ProfileName
225 Uses and attaches ProfileName configuration profile to the vol‐
226 ume group metadata. Whenever the volume group is processed next
227 time, the profile is automatically applied. The profile is
228 inherited by all logical volumes in the volume group unless the
229 logical volume itself has its own profile attached. See
230 lvm.conf(5) for more information about metadata profiles.
231 --[vg]metadatacopies NumberOfCopies|unmanaged|all
233 Sets the desired number of metadata copies in the volume group.
234 If set to a non-zero value, LVM will automatically manage the
235 'metadataignore' flags on the physical volumes (see pvchange or
236 pvcreate --metadataignore) in order to achieve NumberOfCopies
237 copies of metadata. If set to unmanaged, LVM will not automati‐
238 cally manage the 'metadataignore' flags. If set to all, LVM
239 will first clear all of the 'metadataignore' flags on all meta‐
240 data areas in the volume group, then set the value to unmanaged.
241 The vgmetadatacopies option is useful for volume groups contain‐
242 ing large numbers of physical volumes with metadata as it may be
243 used to minimize metadata read and write overhead.
244 -s, --physicalextentsize PhysicalExtentSize[BbBsSkKmMgGtTpPeE]
246 Changes the physical extent size on physical volumes of this
247 volume group. A size suffix (k for kilobytes up to t for ter‐
248 abytes) is optional, megabytes is the default if no suffix is
249 present. For LVM2 format, the value must be a power of 2 of at
250 least 1 sector (where the sector size is the largest sector size
251 of the PVs currently used in the VG) or, if not a power of 2, at
252 least 128KiB. For the older LVM1 format, it must be a power of
253 2 of at least 8KiB. The default is 4 MiB.
254 Before increasing the physical extent size, you might need to
256 use lvresize, pvresize and/or pvmove so that everything fits.
257 For example, every contiguous range of extents used in a logical
258 volume must start and end on an extent boundary.
259 If the volume group metadata uses lvm1 format, extents can vary
261 in size from 8KiB to 16GiB and there is a limit of 65534 extents
262 in each logical volume. The default of 4 MiB leads to a maximum
263 logical volume size of around 256GiB.
264 If the volume group metadata uses lvm2 format those restrictions
266 do not apply, but having a large number of extents will slow
267 down the tools but have no impact on I/O performance to the log‐
268 ical volume. The smallest PE is 1KiB.
269 The 2.4 kernel has a limitation of 2TiB per block device.
271 --systemid SystemID
273 Changes the system ID of the VG. Using this option requires
274 caution because the VG may become foreign to the host running
275 the command, leaving the host unable to access it. See lvmsys‐
276 temid(7).
277 --refresh
279 If any logical volume in the volume group is active, reload its
280 metadata. This is not necessary in normal operation, but may be
281 useful if something has gone wrong or if you're doing clustering
282 manually without a clustered lock manager.
283 -x, --resizeable {y|n}
285 Enables or disables the extension/reduction of this volume group
286 with/by physical volumes.
287
289 To activate all known volume groups in the system:
290 vgchange -a y
292 To change the maximum number of logical volumes of inactive volume
294 group vg00 to 128.
295 vgchange -l 128 /dev/vg00
297
301 lvchange(8), lvm(8), vgcreate(8)
302Sistina Software UKLVM TOOLS 2.02.143(2)-RHEL6 (2016-12-13) VGCHANGE(8)