mount.lustre \- start a Lustre client or target service
.SH SYNOPSIS
.br
-.BI "mount \-t lustre [\-o " options "] " "device mountpoint"
+.BI "mount \-t lustre [\-o " options "] " "<block_device> <server_mountpoint>"
+.br
+.BI "mount \-t lustre [\-o " options "] " "<mgsname>:/<fsname> <client_mountpoint>"
.SH DESCRIPTION
.B mount.lustre
-is used to start a Lustre client or target service. This program should not be
-called directly; rather it is a helper program invoked through
+is used to start a Lustre client or target service. This program should not
+be called directly; rather it is a helper program invoked through
.BR mount (8)
-as above. Lustre clients and targets are stopped by using the
+as above when the
+.B lustre
+filesystem type is used. Lustre clients and targets are stopped by using the
.BR umount (8)
command.
.br
-There are two forms for the
-.I device
-option, depending on whether a client or a target service is started:
+There are two forms for the device option passed to the
+.I mount
+command, depending on whether a client or a target service is started:
.TP
.IR <mgsname> :/ <fsname>[/<subdir>]
mounts the Lustre filesystem named
(optionally starting at subdirectory
.I subdir
within the filesystem, if specified) on the client at the directory
-.IR mountpoint ,
+.IR client_mountpoint ,
by contacting the Management Service at
.IR mgsname .
The format for
.I mgsname
-is defined below. A client filesystem can be listed in
+is defined below. A client filesystem will typically be listed in
.BR fstab (5)
for automatic mount at boot time, is usable like any local filesystem, and
provides a full POSIX-compilant interface.
.I block_device
starts the target service defined by the
.IR mkfs.lustre (8)
-command on the physical disk
+command on the physical disk or ZFS dataset given by
.IR block_device .
The
.I block_device
.BI <mgsnode>:= <mgsnid>[,<mgsnid>]
Each
.I mgsnode
-may be specify a comma-separated list of NIDs, if there are different
-LNet interfaces for that
+may specify a comma-separated list of NIDs, if there are different
+LNet interfaces for the same
.IR mgsnode .
.TP
.BI mgssec= flavor
.BR flock(2)
will get an
.B ENOSYS
-error. It is up to theadministrator to choose either the
+error. It is up to the administrator to choose either the
.B localflock
or
.B flock
Force a client to keep pinging even if servers have enabled suppress_pings.
.TP
.BI verbose
-Enable mount/umount console messages.
+Enable mount/remount/umount console messages.
.TP
.BI noverbose
-Disable mount/umount console messages.
+Disable mount/remount/umount console messages.
.TP
.BI user_fid2path
-Enable FID to path translation by regular users.
+Enable FID to path translation by regular users via
+.BR "lfs fid2path" .
.TP
.BI nouser_fid2path
Disable FID to path translation by regular users. Root and process with
points on the same node, with each mount point using a different
network. It is also interesting when running Lustre clients from
containers, by restricting each container to a specific network.
-.SH CLIENT OPTIONS
+.SH SERVER OPTIONS
In addition to the standard mount options and backing disk type
(e.g. ldiskfs) options listed in
.BR mount (8),
options:
.TP
.BI acl
-Enable POSIX Access Control List support. See the
+Enable POSIX Access Control List support for all clients. See the
.BR acl (5)
manual page.
.TP
.BI nosvc
-Only start the MGC (and MGS, if co-located) for a target service, and not the actual service.
+Only start the MGC (and MGS, if co-located) for a target service,
+and not the actual MDS or OSS service.
.TP
.BI nomgs
-Start a MDT with a co-located MGS without starting the MGS.
+Start an MDT with a co-located MGS without starting the MGS service.
.TP
.BI noscrub
-Not trigger OI scrub automatically when detect some inconsistency, unless it is started explicitly.
+Do not trigger OI scrub automatically when an inconsistency is detected.
+It is still possible to start explicitly using the
+.BR "lctl lfsck_start" .
.TP
.BI skip_lfsck
-Not resume the former paused/crashed LFSCK automatically when mount.
+Do not resume the former paused/crashed LFSCK automatically when mounting.
.TP
.BI abort_recov
Abort client recovery and start the target service immediately.
{
fprintf(out, "\nThis mount helper should only be invoked via the "
"mount (8) command,\ne.g. mount -t lustre dev dir\n\n");
- fprintf(out, "usage: %s [-fhnvV] [-o <mntopt>] <device> <mountpt>\n",
+ fprintf(out, "usage: %s [-fhnvV] [-o <srvopt>] <device> <mountpt>\n",
+ progname);
+ fprintf(out, "usage: %s [-fhnvV] [-o <cliopt>] <mgstarget> <mountpt>\n",
progname);
fprintf(out,
- "\t<device>: the disk device, or for a client:\n"
+ "\t<device>: the local disk device when mounting a server\n"
+ "\t<mgstarget>: the server MGS and filesystem for a client:\n"
"\t\t<mgsnid>[:<altmgsnid>...]:/<filesystem>[/<subdir>]\n"
- "\t<filesystem>: name of the Lustre filesystem (e.g. lustre1)\n"
+ "\t\t\t<mgsnid>: MGS LNet Node Identifier (e.g. mgs01@o2ib)\n"
+ "\t\t\t<filesystem>: Lustre filesystem name (e.g. lustre1)\n"
+ "\t\t\t<subdir>: subdirectory of the filesystem to mount\n"
"\t<mountpt>: filesystem mountpoint (e.g. /mnt/lustre)\n"
- "\t-f|--fake: fake mount (updates /etc/mtab)\n"
+ "\t-f|--fake: fake mount (only update /etc/mtab)\n"
"\t-o force|--force: force mount even if already in /etc/mtab\n"
"\t-h|--help: print this usage message\n"
"\t-n|--nomtab: do not update /etc/mtab after mount\n"
"\t-v|--verbose: print verbose config settings\n"
"\t-V|--version: output build version of the utility and exit\n"
- "\t<mntopt>: one or more comma separated of:\n"
- "\t\t(no)flock,(no)user_xattr,(no)acl\n"
+ "\tdefault options are marked below with '*'\n"
+ "\t\t(no)flock: disable* or enable POSIX flock support\n"
+ "\t\t(no)user_xattr: disable or enable* user xattr namespace\n"
+ "\t<srvopt>: one or more comma separated server options:\n"
+ "\t\t(no)acl: disable or enable* POSIX ACL support completely\n"
"\t\tabort_recov: abort server recovery handling\n"
- "\t\tnosvc: only start MGC/MGS obds\n"
- "\t\tnomgs: only start target obds, using existing MGS\n"
- "\t\tnoscrub: NOT auto start OI scrub unless start explicitly\n"
- "\t\tskip_lfsck: NOT auto resume the paused/crashed LFSCK\n"
- "\t\texclude=<ostname>[:<ostname>] : colon-separated list of "
- "inactive OSTs (e.g. lustre-OST0001)\n"
+ "\t\tnosvc: only start MGC/MGS without starting MDS/OSS\n"
+ "\t\tnomgs: only start target MDS/OSS, using existing MGS\n"
+ "\t\tnoscrub: do NOT auto start OI scrub unless requested\n"
+ "\t\tskip_lfsck: do NOT auto resume paused/crashed LFSCK\n"
+ "\t\tmd_stripe_cache_size=<num>: set MD RAID device stripe cache size\n"
+ "\t<cliopt>: one or more comma separated client options:\n"
+ "\t\texclude=<ostname>[:<ostname>]: list of inactive OSTs (e.g. lustre-OST0001)\n"
+ "\t\tlocalflock: enable POSIX flock only on local client\n"
+ "\t\tretry=<num>: number of times mount is retried by client\n"
#ifdef HAVE_GSS
- "\t\tskpath=<file|directory>: Path to a file or directory of"
- "key configuration files to load into the kernel keyring\n"
+ "\t\tskpath=<file|directory>: path of keys to load into kernel keyring\n"
#endif
- "\t\tretry=<num>: number of times mount is retried by client\n"
- "\t\tmd_stripe_cache_size=<num>: set the raid stripe cache "
- "size for the underlying raid if present\n");
+ "\t\t(no)user_fid2path: disable* or enable user $MOUNT/.lustre/fid access\n"
+ "\t\t(no)checksum: disable or enable* data checksums\n"
+ "\t\t(no)lruresize: disable or enable* LDLM dynamic LRU size\n"
+ "\t\t(no)lazystatfs: disable or enable* statfs to work if OST is unavailable\n"
+ "\t\t32bitapi: return only 32-bit inode numbers to userspace\n"
+ "\t\t(no)verbose: disable or enable* messages at filesystem (un,re)mount\n"
+ );
exit((out != stdout) ? EINVAL : 0);
}