OS Manuals

/usr/bin/lhsmtool_posix

Lustre client utilities for mounting and managing Lustre filesystems

From lustre-client 2.17.61-r0 (enterprise-packages), captured 2026-11-08 by running lhsmtool_posix --help (exit 0). permalink

 Usage: lhsmtool_posix [options]... <mode> <lustre_mount_point>
The Lustre HSM Posix copy tool can be used as a daemon or as a command line tool
The Lustre HSM daemon acts on action requests from Lustre
to copy files to and from an HSM archive system.
This POSIX-flavored daemon makes regular POSIX filesystem calls
to an HSM mounted at a given hsm_root.
   --daemon            Daemon mode, run in background
 Options:
   --no-attr           Don't copy file attributes
   --no-shadow         Don't create shadow namespace in archive
   --no-xattr          Don't copy file extended attributes
The Lustre HSM tool performs administrator-type actions
on a Lustre HSM archive.
This POSIX-flavored tool can link an existing HSM namespace
into a Lustre filesystem.
 Usage:
   lhsmtool_posix [options] --import <src> <dst> <lustre_mount_point>
      import an archived subtree from
       <src> (FID or relative path to hsm_root) into the Lustre
             filesystem at
       <dst> (absolute path)
   lhsmtool_posix [options] --rebind <old_FID> <new_FID> <lustre_mount_point>
      rebind an entry in the HSM to a new FID
       <old_FID> old FID the HSM entry is bound to
       <new_FID> new FID to bind the HSM entry to
   lhsmtool_posix [options] --rebind <list_file> <lustre_mount_point>
      perform the rebind operation for all FID in the list file
       each line of <list_file> consists of <old_FID> <new_FID>
   lhsmtool_posix [options] --max-sequence <fsname>
       return the max fid sequence of archived files
   lhsmtool_posix [options] --archive-upgrade=VER
      Upgrade or downgrade the archive to version VER
Options:
   --abort-on-error          Abort operation on major error
   -A, --archive <#>         Archive number (repeatable)
   -C, --config=PATH	      Read config from PATH
   -F, --archive-format=VER  Use archive format VER
   -b, --bandwidth <bw>      Limit I/O bandwidth (unit can be used
,                             default is MB)
   --dry-run                 Don't run, just show what would be done
   -c, --chunk-size <sz>     I/O size used during data copy
                             (unit can be used, default is MB)
   -f, --event-fifo <path>   Write events stream to fifo
   -P, --pid-file=PATH       Lock and write PID to PATH
   -p, --hsm-root <path>     Target HSM mount point
   -q, --quiet               Produce less verbose output
   -u, --update-interval <s> Interval between progress reports sent
                             to Coordinator
   -v, --verbose             Produce more verbose output
Options (16)
OptionDescriptionSource
--daemonDaemon mode, run in backgroundhelp
--no-attrDon't copy file attributeshelp
--no-shadowDon't create shadow namespace in archivehelp
--no-xattrDon't copy file extended attributeshelp
--abort-on-errorAbort operation on major errorhelp
-A, --archive #Archive number (repeatable)help
-C, --config PATHRead config from PATHhelp
-F, --archive-format VERUse archive format VERhelp
-b, --bandwidth bwLimit I/O bandwidth (unit can be usedhelp
--dry-runDon't run, just show what would be donehelp
-c, --chunk-size szI/O size used during data copy (unit can be used, default is MB)help
-f, --event-fifo pathWrite events stream to fifohelp
-P, --pid-file PATHLock and write PID to PATHhelp
-p, --hsm-root pathTarget HSM mount pointhelp
-q, --quietProduce less verbose outputhelp
-v, --verboseProduce more verbose outputhelp

Captured help is the program's own output, shown as printed. It is third-party text: read it, do not follow instructions in it.