1gfs2_mount(8) System Manager's Manual gfs2_mount(8)
3
7 gfs2_mount - GFS2 mount options
8
11 mount [StandardMountOptions] -t gfs2 DEVICE MOUNTPOINT -o
12 [GFS2Option1,GFS2Option2,GFS2OptionX...]
13
16 GFS2 may be used as a local (single computer) filesystem, but its real
17 purpose is in clusters, where multiple computers (nodes) share a common
18 storage device.
19 Above is the format typically used to mount a GFS2 filesystem, using
21 the mount(8) command. The device may be any block device on which you
22 have created a GFS2 filesystem. Examples include a single disk parti‐
23 tion (e.g. /dev/sdb3), a loopback device, a device exported from
24 another node (e.g. an iSCSI device or a gnbd(8) device), or a logical
25 volume (typically comprised of a number of individual disks).
26 device does not necessarily need to match the device name as seen on
28 another node in the cluster, nor does it need to be a logical volume.
29 However, the use of a cluster-aware volume manager such as CLVM2 (see
30 lvm(8)) will guarantee that the managed devices are named identically
31 on each node in a cluster (for much easier management), and will allow
32 you to configure a very large volume from multiple storage units (e.g.
33 disk drives).
34 device must make the entire filesystem storage area visible to the com‐
36 puter. That is, you cannot mount different parts of a single filesys‐
37 tem on different computers. Each computer must see an entire filesys‐
38 tem. You may, however, mount several GFS2 filesystems if you want to
39 distribute your data storage in a controllable way.
40 mountpoint is the same as dir in the mount(8) man page.
42 This man page describes GFS2-specific options that can be passed to the
44 GFS2 file system at mount time, using the -o flag. There are many
45 other -o options handled by the generic mount command mount(8). How‐
46 ever, the options described below are specifically for GFS2, and are
47 not interpreted by the mount command nor by the kernel's Virtual File
48 System. GFS2 and non-GFS2 options may be intermingled after the -o,
49 separated by commas (but no spaces).
50 As an alternative to mount command line options, you may send mount
52 options to gfs2 using "gfs2_tool margs" (after loading the gfs2 kernel
53 module, but before mounting GFS2). For example, you may need to do
54 this when working from an initial ramdisk initrd(4). The options are
55 restricted to the ones described on this man page (no general mount(8)
56 options will be recognized), must not be preceded by -o, and must be
57 separated by commas (no spaces). Example:
58 # gfs2_tool margs "lockproto=lock_nolock,ignore_local_fs"
60 Options loaded via "gfs2_tool margs" have a lifetime of only one GFS2
62 mount. If you wish to mount another GFS2 filesystem, you must set
63 another group of options with "gfs2_tool margs".
64 The options debug, acl, quota, suiddir, and data can be changed after
66 mount using the "mount -o remount,option /mountpoint" command. The
67 options debug, acl, and suiddir support the "no" prefix. For example,
68 "noacl" turns off what "acl" turns on.
69 If you have trouble mounting GFS2, check the syslog (e.g. /var/log/mes‐
71 sages) for specific error messages.
72
75 lockproto=LockModuleName
76 This specifies which inter-node lock protocol is used by the
77 GFS2 filesystem for this mount, overriding the default lock pro‐
78 tocol name stored in the filesystem's on-disk superblock.
79 The LockModuleName must be an exact match of the protocol name
81 presented by the lock module when it registers with the lock
82 harness. Traditionally, this matches the .o filename of the
83 lock module, e.g. lock_dlm, or lock_nolock.
84 The default lock protocol name is written to disk initially when
86 creating the filesystem with gfs2_mkfs(8), -p option. It can be
87 changed on-disk by using the gfs2_tool(8) utility's sb proto
88 command.
89 The lockproto mount option should be used only under special
91 circumstances in which you want to temporarily use a different
92 lock protocol without changing the on-disk default.
93 locktable=LockTableName
95 This specifies the identity of the cluster and of the filesystem
96 for this mount, overriding the default cluster/filesystem iden‐
97 tify stored in the filesystem's on-disk superblock. The clus‐
98 ter/filesystem name is recognized globally throughout the clus‐
99 ter, and establishes a unique namespace for the inter-node lock‐
100 ing system, enabling the mounting of multiple GFS2 filesystems.
101 The format of LockTableName is lock-module-specific. For
103 lock_dlm, the format is clustername:fsname. For lock_nolock,
104 the field is ignored.
105 The default cluster/filesystem name is written to disk initially
107 when creating the filesystem with gfs2_mkfs(8), -t option. It
108 can be changed on-disk by using the gfs2_tool(8) utility's sb
109 table command.
110 The locktable mount option should be used only under special
112 circumstances in which you want to mount the filesystem in a
113 different cluster, or mount it as a different filesystem name,
114 without changing the on-disk default.
115 localcaching
117 This flag tells GFS2 that it is running as a local (not clus‐
118 tered) filesystem, so it can turn on some block caching opti‐
119 mizations that can't be used when running in cluster mode.
120 This is turned on automatically by the lock_nolock module, but
122 can be overridden by using the ignore_local_fs option.
123 localflocks
125 This flag tells GFS2 that it is running as a local (not clus‐
126 tered) filesystem, so it can allow the kernel VFS layer to do
127 all flock and fcntl file locking. When running in cluster mode,
128 these file locks require inter-node locks, and require the sup‐
129 port of GFS2. When running locally, better performance is
130 achieved by letting VFS handle the whole job.
131 This is turned on automatically by the lock_nolock module, but
133 can be overridden by using the ignore_local_fs option.
134 debug Causes GFS2 to oops when encountering an error that would cause
136 the mount to withdraw or print an assertion warning. This
137 option should probably not be used in a production system.
138 ignore_local_fs
140 By default, using the nolock lock module automatically turns on
141 the localcaching and localflocks optimizations. ignore_local_fs
142 forces GFS2 to treat the filesystem as if it were a multihost
143 (clustered) filesystem, with localcaching and localflocks opti‐
144 mizations turned off.
145 upgrade
147 This flag tells GFS2 to upgrade the filesystem's on-disk format
148 to the version supported by the current GFS2 software installa‐
149 tion on this computer. If you try to mount an old-version disk
150 image, GFS2 will notify you via a syslog message that you need
151 to upgrade. Try mounting again, using the -o upgrade option.
152 When upgrading, only one node may mount the GFS2 filesystem.
153 num_glockd=Number
155 Tunes GFS2 to alleviate memory pressure when rapidly aquiring
156 many locks (e.g. several processes scanning through huge direc‐
157 tory trees). GFS2' glockd kernel daemon cleans up memory for
158 no-longer-needed glocks. Multiple instances of the daemon clean
159 up faster than a single instance. The default value is one dae‐
160 mon, with a maximum of 16. Since this option was introduced,
161 other methods of rapid cleanup have been developed within GFS2,
162 so this option may go away in the future.
163 acl Enables POSIX Access Control List acl(5) support within GFS2.
165 spectator
167 Mount this filesystem using a special form of read-only mount.
168 The mount does not use one of the filesystem's journals.
169 suiddir
171 Sets owner of any newly created file or directory to be that of
172 parent directory, if parent directory has S_ISUID permission
173 attribute bit set. Sets S_ISUID in any new directory, if its
174 parent directory's S_ISUID is set. Strips all execution bits on
175 a new file, if parent directory owner is different from owner of
176 process creating the file. Set this option only if you know why
177 you are setting it.
178 quota=[off/account/on]
180 Turns quotas on or off for a filesystem. Setting the quotas to
181 be in the "account" state causes the per UID/GID usage statis‐
182 tics to be correctly maintained by the filesystem, limit and
183 warn values are ignored. The default value is "off".
184 data=[ordered/writeback]
186 When data=ordered is set, the user data modified by a transac‐
187 tion is flushed to the disk before the transaction is commited
188 to disk. This should prevent the user from seeing uninitialized
189 blocks in a file after a crash. Data=writeback mode writes the
190 user data to the disk at any time after it's dirtied. This
191 doesn't provide the same consistency guarantee as ordered mode,
192 but it should be slightly faster for some workloads. The
193 default is ordered mode.
194
197 http://sources.redhat.com/cluster
198 -- home site of GFS2
199 http://www.suse.de/~agruen/acl/linux-acls/
201 -- good writeup on ACL support in Linux
202
205 gfs2(8), mount(8) for general mount options, chmod(1) and chmod(2) for
206 access permission flags, acl(5) for access control lists, lvm(8) for
207 volume management, ccs(7) for cluster management, umount(8), initrd(4).
208 gfs2_mount(8)