UserUtilities.xml

   1 <?xml version="1.0" encoding="utf-8"?>
   2 <chapter xmlns="http://docbook.org/ns/docbook"
   3 xmlns:xl="http://www.w3.org/1999/xlink" version="5.0" xml:lang="en-US"
   4 xml:id="userutilities">
   5   <title xml:id="userutilities.title">User Utilities</title>
   6   <para>This chapter describes user utilities.</para>
   7   <section xml:id="dbdoclet.50438206_94597">
   8     <title>
   9       <indexterm>
  10         <primary>lfs</primary>
  11       </indexterm>
  12       <literal>lfs</literal>
  13     </title>
  14     <para>The
  15     <literal>lfs</literal> utility can be used for user configuration routines
  16     and monitoring.</para>
  17     <section remap="h5">
  18       <title>Synopsis</title>
  19       <screen>
  20 lfs
  21 lfs changelog [--follow] <replaceable>mdt_name</replaceable> [startrec [endrec]]
  22 lfs changelog_clear <replaceable>mdt_name id endrec</replaceable>
  23 lfs check <replaceable>mds|osts|servers</replaceable>
  24 lfs df [-i] [-h] [--pool]-p <replaceable>fsname</replaceable>[.<replaceable>pool</replaceable>] [<replaceable>path</replaceable>] [--lazy]
  25 lfs find [[!] --atime|-A [-+]N] [[!] --mtime|-M [-+]N]
  26          [[!] --ctime|-C [-+]N] [--maxdepth|-D N] [--name|-n <replaceable>pattern</replaceable>]
  27          [--print|-p] [--print0|-P] [[!] --obd|-O <replaceable>ost_name</replaceable>[,<replaceable>ost_name...</replaceable>]]
  28          [[!] --size|-S [+-]N[kMGTPE]] --type |-t {bcdflpsD}]
  29          [[!] --gid|-g|--group|-G <replaceable>gname|gid</replaceable>]
  30          [[!] --uid|-u|--user|-U <replaceable>uname|uid</replaceable>]
  31          <replaceable>dirname|filename</replaceable>
  32 lfs getname [-h]|[path...]
  33 lfs getstripe [--obd|-O <replaceable>ost_name</replaceable>] [--quiet|-q] [--verbose|-v]
  34               [--count|-c] [--index|-i | --offset|-o]
  35               [--size|-s] [--pool|-p] [--directory|-d]
  36               [--recursive|-r] [--raw|-R] [-M]
  37               <replaceable>dirname|filename</replaceable> ...
  38 lfs setstripe [--size|-s stripe_size] [--count|-c <replaceable>stripe_count</replaceable>]
  39               [--index|-i|--offset|-o <replaceable>start_ost_index</replaceable>]
  40               [--pool|-p <replaceable>pool</replaceable>]
  41               <replaceable>dirname|filename</replaceable>
  42 lfs setstripe -d <replaceable>dir</replaceable>
  43 lfs osts [path]
  44 lfs poollist <replaceable>filesystem</replaceable>[.<replaceable>pool</replaceable>]| <replaceable>pathname</replaceable>
  45 lfs quota [-q] [-v] [-h] [-o <replaceable>obd_uuid</replaceable>|-I <replaceable>ost_idx</replaceable>|-i <replaceable>mdt_idx</replaceable>]
  46           [-u <replaceable>username|uid|-g</replaceable> <replaceable>group|gid</replaceable>] <replaceable>/mount_point</replaceable>
  47 lfs quota -t -u|-g <replaceable>/mount_point</replaceable>
  48 lfs quotacheck [-ug] <replaceable>/mount_point</replaceable>
  49 lfs quotachown [-i] <replaceable>/mount_point</replaceable>
  50 lfs quotainv [-ug] [-f] <replaceable>/mount_point</replaceable>
  51 lfs quotaon [-ugf] <replaceable>/mount_point</replaceable>
  52 lfs quotaoff [-ug] <replaceable>/mount_point</replaceable>
  53 lfs setquota {-u|--user|-g|--group} <replaceable>uname|uid|gname|gid</replaceable>
  54              [--block-softlimit <replaceable>block_softlimit</replaceable>]
  55              [--block-hardlimit <replaceable>block_hardlimit</replaceable>]
  56              [--inode-softlimit <replaceable>inode_softlimit</replaceable>]
  57              [--inode-hardlimit <replaceable>inode_hardlimit</replaceable>]
  58              <replaceable>/mount_point</replaceable>
  59 lfs setquota -u|--user|-g|--group <replaceable>uname|uid|gname|gid</replaceable>
  60              [-b <replaceable>block_softlimit</replaceable>] [-B <replaceable>block_hardlimit</replaceable>]
  61              [-i <replaceable>inode-softlimit</replaceable>] [-I <replaceable>inode_hardlimit</replaceable>]
  62              <replaceable>/mount_point</replaceable>
  63 lfs setquota -t -u|-g [--block-grace <replaceable>block_grace</replaceable>]
  64              [--inode-grace <replaceable>inode_grace</replaceable>]
  65              <replaceable>/mount_point</replaceable>
  66 lfs setquota -t -u|-g [-b <replaceable>block_grace</replaceable>] [-i <replaceable>inode_grace</replaceable>]
  67              <replaceable>/mount_point</replaceable>
  68 lfs help
  69 </screen>
  70       <note>
  71         <para>In the above example, the
  72         <literal>
  73           <replaceable>/mount_point</replaceable>
  74         </literal> parameter refers to the mount point of the Lustre file
  75         system.</para>
  76       </note>
  77       <note>
  78         <para>The old lfs quota output was very detailed and contained
  79         cluster-wide quota statistics (including cluster-wide limits for a
  80         user/group and cluster-wide usage for a user/group), as well as
  81         statistics for each MDS/OST. Now,
  82         <literal>lfs quota</literal> has been updated to provide only
  83         cluster-wide statistics, by default. To obtain the full report of
  84         cluster-wide limits, usage and statistics, use the
  85         <literal>-v</literal> option with
  86         <literal>lfs quota</literal>.</para>
  87       </note>
  88     </section>
  89     <section remap="h5">
  90       <title>Description</title>
  91       <para>The
  92       <literal>lfs</literal> utility is used to create a new file with a
  93       specific striping pattern, determine the default striping pattern, gather
  94       the extended attributes (object numbers and location) for a specific
  95       file, find files with specific attributes, list OST information or set
  96       quota limits. It can be invoked interactively without any arguments or in
  97       a non-interactive mode with one of the supported arguments.</para>
  98     </section>
  99     <section remap="h5">
 100       <title>Options</title>
 101       <para>The various
 102       <literal>lfs</literal> options are listed and described below. For a
 103       complete list of available options, type help at the
 104       <literal>lfs</literal> prompt.</para>
 105       <informaltable frame="all">
 106         <tgroup cols="3">
 107           <colspec colname="c1" colwidth="20*" />
 108           <colspec colname="c2" colwidth="30*" />
 109           <colspec colname="c3" colwidth="50*" />
 110           <thead>
 111             <row>
 112               <entry nameend="c2" namest="c1">
 113                 <para>
 114                   <emphasis role="bold">Option</emphasis>
 115                 </para>
 116               </entry>
 117               <entry>
 118                 <para>
 119                   <emphasis role="bold">Description</emphasis>
 120                 </para>
 121               </entry>
 122             </row>
 123           </thead>
 124           <tbody>
 125             <row>
 126               <entry nameend="c2" namest="c1">
 127                 <para>
 128                   <literal>changelog</literal>
 129                 </para>
 130               </entry>
 131               <entry>
 132                 <para>Shows the metadata changes on an MDT. Start and end
 133                 points are optional. The
 134                 <literal>--follow</literal> option blocks on new changes; this
 135                 option is only valid when run directly on the MDT node.</para>
 136               </entry>
 137             </row>
 138             <row>
 139               <entry nameend="c2" namest="c1">
 140                 <para>
 141                   <literal>changelog_clear</literal>
 142                 </para>
 143               </entry>
 144               <entry>
 145                 <para>Indicates that changelog records previous to
 146                 <literal>
 147                   <replaceable>endrec</replaceable>
 148                 </literal> are no longer of interest to a particular consumer
 149                 <literal>
 150                   <replaceable>id</replaceable>
 151                 </literal>, potentially allowing the MDT to free up disk space.
 152                 An
 153                 <literal>
 154                   <replaceable>endrec</replaceable>
 155                 </literal> of 0 indicates the current last record. Changelog
 156                 consumers must be registered on the MDT node using
 157                 <literal>lctl</literal>.</para>
 158               </entry>
 159             </row>
 160             <row>
 161               <entry nameend="c2" namest="c1">
 162                 <literal>check</literal>
 163               </entry>
 164               <entry>
 165                 <para>Displays the status of MDS or OSTs (as specified in the
 166                 command) or all servers (MDS and OSTs).</para>
 167               </entry>
 168             </row>
 169             <row>
 170               <entry nameend="c2" namest="c1">
 171                 <literal>df [-i] [-h] [--pool|-p
 172                 <replaceable>fsname</replaceable>[.
 173                 <replaceable>pool</replaceable>] [
 174                 <replaceable>path</replaceable>] [--lazy]</literal>
 175               </entry>
 176               <entry>
 177                 <para>Use
 178                 <literal>-i</literal> to report file system disk space usage or
 179                 inode usage of each MDT or OST or, if a pool is specified with
 180                 the
 181                 <literal>-p</literal> option, a subset of OSTs.</para>
 182                 <para>By default, the usage of all mounted Lustre file systems
 183                 is reported. If the
 184                 <literal>path</literal> option is included, only the usage for
 185                 the specified file system is reported. If the
 186                 <literal>-h</literal> option is included, the output is printed
 187                 in human-readable format, using SI base-2 suffixes for
 188                 <emphasis role="bold">M</emphasis>ega-,
 189                 <emphasis role="bold">G</emphasis>iga-,
 190                 <emphasis role="bold">T</emphasis>era-,
 191                 <emphasis role="bold">P</emphasis>eta-, or
 192                 <emphasis role="bold">E</emphasis>xabytes.</para>
 193                 <para>If the
 194                 <literal>--lazy</literal> option is specified, any OST that is
 195                 currently disconnected from the client will be skipped. Using
 196                 the
 197                 <literal>--lazy</literal> option prevents the
 198                 <literal>df</literal> output from being blocked when an OST is
 199                 offline. Only the space on the OSTs that can currently be
 200                 accessed are returned. The
 201                 <literal>llite.*.lazystatfs</literal> tunable can be enabled to
 202                 make this the default behaviour for all
 203                 <literal>statfs()</literal> operations.</para>
 204               </entry>
 205             </row>
 206             <row>
 207               <entry nameend="c2" namest="c1">
 208                 <para>
 209                   <literal>find</literal>
 210                 </para>
 211               </entry>
 212               <entry>
 213                 <para>Searches the directory tree rooted at the given
 214                 directory/filename for files that match the given
 215                 parameters.</para>
 216                 <para>Using
 217                 <literal>!</literal> before an option negates its meaning (files
 218                 NOT matching the parameter). Using
 219                 <literal>+</literal> before a numeric value means files with the
 220                 parameter OR MORE. Using
 221                 <literal>-</literal> before a numeric value means files with the
 222                 parameter OR LESS.</para>
 223               </entry>
 224             </row>
 225             <row>
 226               <entry />
 227               <entry>
 228                 <literal>--atime</literal>
 229               </entry>
 230               <entry>
 231                 <para>File was last accessed N*24 hours ago. (There is no
 232                 guarantee that
 233                 <literal>atime</literal> is kept coherent across the
 234                 cluster.)</para>
 235                 <para>OSTs store a transient
 236                 <literal>atime</literal> that is updated when clients do read
 237                 requests. Permanent
 238                 <literal>atime</literal> is written to the MDS when the file is
 239                 closed. However, on-disk atime is only updated if it is more
 240                 than 60 seconds old (
 241                 <literal>/proc/fs/lustre/mds/*/max_atime_diff</literal>). The
 242                 Lustre software considers the latest
 243                 <literal>atime</literal> from all OSTs. If a
 244                 <literal>setattr</literal> is set by user, then it is updated on
 245                 both the MDS and OST, allowing the
 246                 <literal>atime</literal> to go backward.</para>
 247               </entry>
 248             </row>
 249             <row>
 250               <entry>
 251                 <para>&#160;</para>
 252               </entry>
 253               <entry>
 254                 <para>
 255                   <literal>--ctime</literal>
 256                 </para>
 257               </entry>
 258               <entry>
 259                 <para>File status was last changed N*24 hours ago.</para>
 260               </entry>
 261             </row>
 262             <row>
 263               <entry>
 264                 <para>&#160;</para>
 265               </entry>
 266               <entry>
 267                 <para>
 268                   <literal>--mtime</literal>
 269                 </para>
 270               </entry>
 271               <entry>
 272                 <para>File data was last modified N*24 hours ago.</para>
 273               </entry>
 274             </row>
 275             <row>
 276               <entry>
 277                 <para>&#160;</para>
 278               </entry>
 279               <entry>
 280                 <para>
 281                   <literal>--obd</literal>
 282                 </para>
 283               </entry>
 284               <entry>
 285                 <para>File has an object on a specific OST(s).</para>
 286               </entry>
 287             </row>
 288             <row>
 289               <entry>
 290                 <para>&#160;</para>
 291               </entry>
 292               <entry>
 293                 <para>
 294                   <literal>--size</literal>
 295                 </para>
 296               </entry>
 297               <entry>
 298                 <para>File has a size in bytes, or kilo-, Mega-, Giga-, Tera-,
 299                 Peta- or Exabytes if a suffix is given.</para>
 300               </entry>
 301             </row>
 302             <row>
 303               <entry>
 304                 <para>&#160;</para>
 305               </entry>
 306               <entry>
 307                 <para>
 308                   <literal>--type</literal>
 309                 </para>
 310               </entry>
 311               <entry>
 312                 <para>File has the type - block, character, directory, pipe,
 313                 file, symlink, socket or door (used in Solaris operating
 314                 system).</para>
 315               </entry>
 316             </row>
 317             <row>
 318               <entry>
 319                 <para>&#160;</para>
 320               </entry>
 321               <entry>
 322                 <para>
 323                   <literal>--uid</literal>
 324                 </para>
 325               </entry>
 326               <entry>
 327                 <para>File has a specific numeric user ID.</para>
 328               </entry>
 329             </row>
 330             <row>
 331               <entry>
 332                 <para>&#160;</para>
 333               </entry>
 334               <entry>
 335                 <para>
 336                   <literal>--user</literal>
 337                 </para>
 338               </entry>
 339               <entry>
 340                 <para>File owned by a specific user (numeric user ID
 341                 allowed).</para>
 342               </entry>
 343             </row>
 344             <row>
 345               <entry>
 346                 <para>&#160;</para>
 347               </entry>
 348               <entry>
 349                 <para>
 350                   <literal>--gid</literal>
 351                 </para>
 352               </entry>
 353               <entry>
 354                 <para>File has a specific group ID.</para>
 355               </entry>
 356             </row>
 357             <row>
 358               <entry>
 359                 <para>&#160;</para>
 360               </entry>
 361               <entry>
 362                 <para>
 363                   <literal>--group</literal>
 364                 </para>
 365               </entry>
 366               <entry>
 367                 <para>File belongs to a specific group (numeric group ID
 368                 allowed).</para>
 369               </entry>
 370             </row>
 371             <row>
 372               <entry>
 373                 <para>&#160;</para>
 374               </entry>
 375               <entry>
 376                 <para>-
 377                 <literal>-maxdepth</literal></para>
 378               </entry>
 379               <entry>
 380                 <para>Limits find to descend at most N levels of the directory
 381                 tree.</para>
 382               </entry>
 383             </row>
 384             <row>
 385               <entry>
 386                 <para>&#160;</para>
 387               </entry>
 388               <entry>
 389                 <para>
 390                 <literal>--print</literal>/
 391                 <literal>--print0</literal></para>
 392               </entry>
 393               <entry>
 394                 <para>Prints the full filename, followed by a new line or NULL
 395                 character correspondingly.</para>
 396               </entry>
 397             </row>
 398             <row>
 399               <entry nameend="c2" namest="c1">
 400                 <para>
 401                   <literal>osts [path]</literal>
 402                 </para>
 403               </entry>
 404               <entry>
 405                 <para>Lists all OSTs for the file system. If a path located on
 406                 a mounted Lustre file system is specified, then only OSTs
 407                 belonging to this file system are displayed.</para>
 408               </entry>
 409             </row>
 410             <row>
 411               <entry nameend="c2" namest="c1">
 412                 <para>
 413                   <literal>getname [path...]</literal>
 414                 </para>
 415               </entry>
 416               <entry>
 417                 <para>List each Lustre file system instance associated with
 418                 each Lustre mount point. If no path is specified, all Lustre
 419                 mount points are interrogated. If a list of paths is provided,
 420                 the instance of each path is provided. If the path is not a
 421                 Lustre instance 'No such device' is returned.</para>
 422               </entry>
 423             </row>
 424             <row>
 425               <entry nameend="c2" namest="c1">
 426                 <para>
 427                   <literal>getstripe</literal>
 428                 </para>
 429               </entry>
 430               <entry>
 431                 <para>Lists striping information for a given filename or
 432                 directory. By default, the stripe count, stripe size and offset
 433                 are returned.</para>
 434                 <para>If you only want specific striping information, then the
 435                 options of
 436                 <literal>--count</literal>,
 437                 <literal>--size</literal>,
 438                 <literal>--index</literal> or
 439                 <literal>--offset</literal> plus various combinations of these
 440                 options can be used to retrieve specific information.</para>
 441                 <para>If the
 442                 <literal>--raw</literal> option is specified, the stripe
 443                 information is printed without substituting the file system
 444                 default values for unspecified fields. If the striping EA is
 445                 not set, 0, 0, and -1 will be printed for the stripe count,
 446                 size, and offset respectively.</para>
 447                 <para condition="l24">The
 448                 <literal>-M</literal> prints the index of the MDT for a given
 449                 directory. See
 450                 <xref linkend="dbdoclet.rmremotedir" />.</para>
 451               </entry>
 452             </row>
 453             <row>
 454               <entry>
 455                 <para>&#160;</para>
 456               </entry>
 457               <entry>
 458                 <para>
 459                   <literal>--obd
 460                   <replaceable>ost_name</replaceable></literal>
 461                 </para>
 462               </entry>
 463               <entry>
 464                 <para>Lists files that have an object on a specific OST.</para>
 465               </entry>
 466             </row>
 467             <row>
 468               <entry>
 469                 <para>&#160;</para>
 470               </entry>
 471               <entry>
 472                 <para>
 473                   <literal>--quiet</literal>
 474                 </para>
 475               </entry>
 476               <entry>
 477                 <para>Lists details about the file's object ID
 478                 information.</para>
 479               </entry>
 480             </row>
 481             <row>
 482               <entry>
 483                 <para>&#160;</para>
 484               </entry>
 485               <entry>
 486                 <para>
 487                   <literal>--verbose</literal>
 488                 </para>
 489               </entry>
 490               <entry>
 491                 <para>Prints additional striping information.</para>
 492               </entry>
 493             </row>
 494             <row>
 495               <entry>
 496                 <para>&#160;</para>
 497               </entry>
 498               <entry>
 499                 <para>
 500                   <literal>--count</literal>
 501                 </para>
 502               </entry>
 503               <entry>
 504                 <para>Lists the stripe count (how many OSTs to use).</para>
 505               </entry>
 506             </row>
 507             <row>
 508               <entry>
 509                 <para>&#160;</para>
 510               </entry>
 511               <entry>
 512                 <para>
 513                   <literal>--index</literal>
 514                 </para>
 515               </entry>
 516               <entry>
 517                 <para>Lists the index for each OST in the file system.</para>
 518               </entry>
 519             </row>
 520             <row>
 521               <entry>
 522                 <para>&#160;</para>
 523               </entry>
 524               <entry>
 525                 <para>
 526                   <literal>--offset</literal>
 527                 </para>
 528               </entry>
 529               <entry>
 530                 <para>Lists the OST index on which file striping starts.</para>
 531               </entry>
 532             </row>
 533             <row>
 534               <entry>
 535                 <para>&#160;</para>
 536               </entry>
 537               <entry>
 538                 <para>
 539                   <literal>--pool</literal>
 540                 </para>
 541               </entry>
 542               <entry>
 543                 <para>Lists the pools to which a file belongs.</para>
 544               </entry>
 545             </row>
 546             <row>
 547               <entry>
 548                 <para>&#160;</para>
 549               </entry>
 550               <entry>
 551                 <para>
 552                   <literal>--size</literal>
 553                 </para>
 554               </entry>
 555               <entry>
 556                 <para>Lists the stripe size (how much data to write to one OST
 557                 before moving to the next OST).</para>
 558               </entry>
 559             </row>
 560             <row>
 561               <entry>
 562                 <para>&#160;</para>
 563               </entry>
 564               <entry>
 565                 <para>
 566                   <literal>--directory</literal>
 567                 </para>
 568               </entry>
 569               <entry>
 570                 <para>Lists entries about a specified directory instead of its
 571                 contents (in the same manner as
 572                 <literal>ls -d</literal>).</para>
 573               </entry>
 574             </row>
 575             <row>
 576               <entry>
 577                 <para>&#160;</para>
 578               </entry>
 579               <entry>
 580                 <para>
 581                   <literal>--recursive</literal>
 582                 </para>
 583               </entry>
 584               <entry>
 585                 <para>Recurses into all sub-directories.</para>
 586               </entry>
 587             </row>
 588             <row>
 589               <entry nameend="c2" namest="c1">
 590                 <para>
 591                   <literal>setstripe</literal>
 592                 </para>
 593               </entry>
 594               <entry>
 595                 <para>Create new files with a specific file layout (stripe
 596                 pattern) configuration.
 597                 <footnote>
 598                   <para>The file cannot exist prior to using
 599                   <literal>setstripe</literal>. A directory must exist prior to
 600                   using
 601                   <literal>setstripe</literal>.</para>
 602                 </footnote></para>
 603               </entry>
 604             </row>
 605             <row>
 606               <entry>
 607                 <para>&#160;</para>
 608               </entry>
 609               <entry>
 610                 <para>
 611                   <literal>--count stripe_cnt</literal>
 612                 </para>
 613               </entry>
 614               <entry>
 615                 <para>Number of OSTs over which to stripe a file. A
 616                 <literal>stripe_cnt</literal> of 0 uses the file system-wide
 617                 default stripe count (default is 1). A
 618                 <literal>stripe_cnt</literal> of -1 stripes over all available
 619                 OSTs.</para>
 620               </entry>
 621             </row>
 622             <row>
 623               <entry>
 624                 <para>&#160;</para>
 625               </entry>
 626               <entry>
 627                 <para>
 628                 <literal>--size stripe_size</literal>
 629                 <footnote>
 630                   <para>The default stripe-size is 0. The default start-ost is
 631                   -1. Do NOT confuse them! If you set start-ost to 0, all new
 632                   file creations occur on OST 0 (seldom a good idea).</para>
 633                 </footnote>&#160;</para>
 634               </entry>
 635               <entry>
 636                 <para>Number of bytes to store on an OST before moving to the
 637                 next OST. A stripe_size of 0 uses the file system's default
 638                 stripe size, (default is 1 MB). Can be specified with
 639                 <emphasis role="bold">k</emphasis>(KB),
 640                 <emphasis role="bold">m</emphasis>(MB), or
 641                 <emphasis role="bold">g</emphasis>(GB), respectively.</para>
 642               </entry>
 643             </row>
 644             <row>
 645               <entry>
 646                 <para>&#160;</para>
 647               </entry>
 648               <entry>
 649                 <para>
 650                   <literal>--index --offset start_ost_index</literal>
 651                 </para>
 652               </entry>
 653               <entry>
 654                 <para>The OST index (base 10, starting at 0) on which to start
 655                 striping for this file. A start_ost_index value of -1 allows
 656                 the MDS to choose the starting index. This is the default
 657                 value, and it means that the MDS selects the starting OST as it
 658                 wants. We strongly recommend selecting this default, as it
 659                 allows space and load balancing to be done by the MDS as
 660                 needed. The
 661                 <literal>start_ost_index</literal> value has no relevance on
 662                 whether the MDS will use round-robin or QoS weighted allocation
 663                 for the remaining stripes in the file.</para>
 664               </entry>
 665             </row>
 666             <row>
 667               <entry>
 668                 <para>&#160;</para>
 669               </entry>
 670               <entry>
 671                 <para>
 672                   <literal>--pool
 673                   <replaceable>pool</replaceable></literal>
 674                 </para>
 675               </entry>
 676               <entry>
 677                 <para>Name of the pre-defined pool of OSTs (see
 678                 <xref linkend="dbdoclet.50438219_38274" />) that will be used
 679                 for striping. The
 680                 <literal>stripe_cnt</literal>,
 681                 <literal>stripe_size</literal> and
 682                 <literal>start_ost</literal> values are used as well. The
 683                 start-ost value must be part of the pool or an error is
 684                 returned.</para>
 685               </entry>
 686             </row>
 687             <row>
 688               <entry nameend="c2" namest="c1">
 689                 <para>
 690                   <literal>setstripe -d</literal>
 691                 </para>
 692               </entry>
 693               <entry>
 694                 <para>Deletes default striping on the specified
 695                 directory.</para>
 696               </entry>
 697             </row>
 698             <row>
 699               <entry nameend="c2" namest="c1">
 700                 <para>
 701                   <literal>poollist {filesystem}
 702                   [.poolname]|{pathname}</literal>
 703                 </para>
 704               </entry>
 705               <entry>
 706                 <para>Lists pools in the file system or pathname, or OSTs in
 707                 the file system's pool.</para>
 708               </entry>
 709             </row>
 710             <row>
 711               <entry nameend="c2" namest="c1">
 712                 <para>
 713                   <literal>quota [-q] [-v] [-o
 714                   <replaceable>obd_uuid</replaceable>|-i
 715                   <replaceable>mdt_idx</replaceable>|-I
 716                   <replaceable>ost_idx</replaceable>] [-u|-g
 717                   <replaceable>uname|uid|gname|gid]</replaceable>
 718                   <replaceable>/mount_point</replaceable></literal>
 719                 </para>
 720                 <para>&#160;</para>
 721               </entry>
 722               <entry>
 723                 <para>Displays disk usage and limits, either for the full file
 724                 system or for objects on a specific OBD. A user or group name
 725                 or an ID can be specified. If both user and group are omitted,
 726                 quotas for the current UID/GID are shown. The
 727                 <literal>-q</literal> option disables printing of additional
 728                 descriptions (including column titles). It fills in blank
 729                 spaces in the
 730                 <literal>grace</literal> column with zeros (when there is no
 731                 grace period set), to ensure that the number of columns is
 732                 consistent. The
 733                 <literal>-v</literal> option provides more verbose (per-OBD
 734                 statistics) output.</para>
 735               </entry>
 736             </row>
 737             <row>
 738               <entry nameend="c2" namest="c1">
 739                 <para>
 740                   <literal>quota -t
 741                   <replaceable>-u|-g</replaceable>
 742                   <replaceable>/mount_point</replaceable></literal>
 743                 </para>
 744               </entry>
 745               <entry>
 746                 <para>Displays block and inode grace times for user (
 747                 <literal>-u</literal>) or group (
 748                 <literal>-g</literal>) quotas.</para>
 749               </entry>
 750             </row>
 751             <row>
 752               <entry nameend="c2" namest="c1">
 753                 <para>
 754                   <literal>quotachown</literal>
 755                 </para>
 756               </entry>
 757               <entry>
 758                 <para>Changes the file's owner and group on OSTs of the
 759                 specified file system.</para>
 760               </entry>
 761             </row>
 762             <row>
 763               <entry nameend="c2" namest="c1">
 764                 <para>
 765                   <literal>quotacheck [-ugf]
 766                   <replaceable>/mount_point</replaceable></literal>
 767                 </para>
 768               </entry>
 769               <entry>
 770                 <para>Scans the specified file system for disk usage, and
 771                 creates or updates quota files. Options specify quota for users
 772                 (
 773                 <literal>-u</literal>), groups (
 774                 <literal>-g</literal>), and force (
 775                 <literal>-f</literal>).</para>
 776               </entry>
 777             </row>
 778             <row>
 779               <entry nameend="c2" namest="c1">
 780                 <para>
 781                   <literal>quotaon [-ugf]
 782                   <replaceable>/mount_point</replaceable></literal>
 783                 </para>
 784               </entry>
 785               <entry>
 786                 <para>Turns on file system quotas. Options specify quota for
 787                 users (
 788                 <literal>-u</literal>), groups (
 789                 <literal>-g</literal>), and force (
 790                 <literal>-f</literal>).</para>
 791               </entry>
 792             </row>
 793             <row>
 794               <entry nameend="c2" namest="c1">
 795                 <para>
 796                   <literal>quotaoff [-ugf]
 797                   <replaceable>/mount_point</replaceable></literal>
 798                 </para>
 799               </entry>
 800               <entry>
 801                 <para>Turns off file system quotas. Options specify quota for
 802                 users (
 803                 <literal>-u</literal>), groups (
 804                 <literal>-g</literal>), and force (
 805                 <literal>-f</literal>).</para>
 806               </entry>
 807             </row>
 808             <row>
 809               <entry nameend="c2" namest="c1">
 810                 <para>
 811                   <literal>quotainv [-ug] [-f]
 812                   <replaceable>/mount_point</replaceable></literal>
 813                 </para>
 814               </entry>
 815               <entry>
 816                 <para>Clears quota files (administrative quota files if used
 817                 without
 818                 <literal>-f</literal>, operational quota files otherwise), all
 819                 of their quota entries for users (
 820                 <literal>-u</literal>) or groups (
 821                 <literal>-g</literal>). After running
 822                 <literal>quotainv</literal>, you must run
 823                 <literal>quotacheck</literal> before using quotas.</para>
 824                 <caution>
 825                   <para>Use extreme caution when using this command; its
 826                   results cannot be undone.</para>
 827                 </caution>
 828               </entry>
 829             </row>
 830             <row>
 831               <entry nameend="c2" namest="c1">
 832                 <para>
 833                   <literal>setquota -u|-g
 834                   <replaceable>
 835                   uname|uid|gname|gid}</replaceable>[--block-softlimit
 836                   <replaceable>block_softlimit</replaceable>]
 837                   [--block-hardlimit
 838                   <replaceable>block_hardlimit</replaceable>]
 839                   [--inode-softlimit
 840                   <replaceable>inode_softlimit</replaceable>]
 841                   [--inode-hardlimit
 842                   <replaceable>inode_hardlimit</replaceable>]
 843                   <replaceable>/mount_point</replaceable></literal>
 844                 </para>
 845               </entry>
 846               <entry>
 847                 <para>Sets file system quotas for users or groups. Limits can
 848                 be specified with
 849                 <literal>--{block|inode}-{softlimit|hardlimit}</literal> or
 850                 their short equivalents
 851                 <literal>-b</literal>,
 852                 <literal>-B</literal>,
 853                 <literal>-i</literal>,
 854                 <literal>-I</literal>. Users can set 1, 2, 3 or 4 limits.
 855                 <footnote>
 856                   <para>The old
 857                   <literal>setquota</literal> interface is supported, but it may
 858                   be removed in a future Lustre software release.</para>
 859                 </footnote>Also, limits can be specified with special suffixes,
 860                 -b, -k, -m, -g, -t, and -p to indicate units of 1, 2^10, 2^20,
 861                 2^30, 2^40 and 2^50, respectively. By default, the block limits
 862                 unit is 1 kilobyte (1,024), and block limits are always
 863                 kilobyte-grained (even if specified in bytes). See
 864                 <xref linkend="dbdoclet.50438206_11903" />.</para>
 865               </entry>
 866             </row>
 867             <row>
 868               <entry nameend="c2" namest="c1">
 869                 <para>
 870                   <literal>setquota -t -u|-g [--block-grace
 871                   <replaceable>block_grace</replaceable>] [--inode-grace
 872                   <replaceable>inode_grace</replaceable>]
 873                   <replaceable>/mount_point</replaceable></literal>
 874                 </para>
 875               </entry>
 876               <entry>
 877                 <para>Sets the file system quota grace times for users or
 878                 groups. Grace time is specified in '
 879                 <literal>XXwXXdXXhXXmXXs</literal>' format or as an integer
 880                 seconds value. See
 881                 <xref linkend="dbdoclet.50438206_11903" />.</para>
 882               </entry>
 883             </row>
 884             <row>
 885               <entry nameend="c2" namest="c1">
 886                 <para>
 887                   <literal>help</literal>
 888                 </para>
 889               </entry>
 890               <entry>
 891                 <para>Provides brief help on various
 892                 <literal>lfs</literal> arguments.</para>
 893               </entry>
 894             </row>
 895             <row>
 896               <entry nameend="c2" namest="c1">
 897                 <para>
 898                   <literal>exit/quit</literal>
 899                 </para>
 900               </entry>
 901               <entry>
 902                 <para>Quits the interactive
 903                 <literal>lfs</literal> session.</para>
 904               </entry>
 905             </row>
 906           </tbody>
 907         </tgroup>
 908       </informaltable>
 909     </section>
 910     <section remap="h5">
 911       <title xml:id="dbdoclet.50438206_11903">Examples</title>
 912       <para>Creates a file striped on two OSTs with 128 KB on each
 913       stripe.</para>
 914       <screen>
 915 $ lfs setstripe -s 128k -c 2 /mnt/lustre/file1
 916 </screen>
 917       <para>Deletes a default stripe pattern on a given directory. New files
 918       use the default striping pattern.</para>
 919       <screen>
 920 $ lfs setstripe -d /mnt/lustre/dir
 921 </screen>
 922       <para>Lists the detailed object allocation of a given file.</para>
 923       <screen>
 924 $ lfs getstripe -v /mnt/lustre/file1
 925 </screen>
 926       <para>List all the mounted Lustre file systems and corresponding Lustre
 927       instances.</para>
 928       <screen>
 929 $ lfs getname
 930 </screen>
 931       <para>Efficiently lists all files in a given directory and its
 932       subdirectories.</para>
 933       <screen>
 934 $ lfs find /mnt/lustre
 935 </screen>
 936       <para>Recursively lists all regular files in a given directory more than
 937       30 days old.</para>
 938       <screen>
 939 $ lfs find /mnt/lustre -mtime +30 -type f -print
 940 </screen>
 941       <para>Recursively lists all files in a given directory that have objects
 942       on OST2-UUID. The lfs check servers command checks the status of all
 943       servers (MDT and OSTs).</para>
 944       <screen>
 945 $ lfs find --obd OST2-UUID /mnt/lustre/
 946 </screen>
 947       <para>Lists all OSTs in the file system.</para>
 948       <screen>
 949 $ lfs osts
 950 </screen>
 951       <para>Lists space usage per OST and MDT in human-readable format.</para>
 952       <screen>
 953 $ lfs df -h
 954 </screen>
 955       <para>Lists inode usage per OST and MDT.</para>
 956       <screen>
 957 $ lfs df -i
 958 </screen>
 959       <para>List space or inode usage for a specific OST pool.</para>
 960       <screen>
 961 $ lfs df --pool
 962 <replaceable>filesystem</replaceable>[.
 963 <replaceable>pool</replaceable>] |
 964 <replaceable>pathname</replaceable>
 965 </screen>
 966       <para>List quotas of user 'bob'.</para>
 967       <screen>
 968 $ lfs quota -u bob /mnt/lustre
 969 </screen>
 970       <para>Show grace times for user quotas on
 971       <literal>/mnt/lustre</literal>.</para>
 972       <screen>
 973 $ lfs quota -t -u /mnt/lustre
 974 </screen>
 975       <para>Changes file owner and group.</para>
 976       <screen>
 977 $ lfs quotachown -i /mnt/lustre
 978 </screen>
 979       <para>Checks quotas for user and group. Turns on quotas after making the
 980       check.</para>
 981       <screen>
 982 $ lfs quotacheck -ug /mnt/lustre
 983 </screen>
 984       <para>Turns on quotas of user and group.</para>
 985       <screen>
 986 $ lfs quotaon -ug /mnt/lustre
 987 </screen>
 988       <para>Turns off quotas of user and group.</para>
 989       <screen>
 990 $ lfs quotaoff -ug /mnt/lustre
 991 </screen>
 992       <para>Sets quotas of user 'bob', with a 1 GB block quota hardlimit and a
 993       2 GB block quota softlimit.</para>
 994       <screen>
 995 $ lfs setquota -u bob --block-softlimit 2000000 --block-hardlimit 1000000
 996 /mnt/lustre
 997 </screen>
 998       <para>Sets grace times for user quotas: 1000 seconds for block quotas, 1
 999       week and 4 days for inode quotas.</para>
1000       <screen>
1001 $ lfs setquota -t -u --block-grace 1000 --inode-grace 1w4d /mnt/lustre
1002 </screen>
1003       <para>Checks the status of all servers (MDT, OST)</para>
1004       <screen>
1005 $ lfs check servers
1006 </screen>
1007       <para>Creates a file striped on two OSTs from the pool
1008       <literal>my_pool</literal></para>
1009       <screen>
1010 $ lfs setstripe --pool my_pool -c 2 /mnt/lustre/file
1011 </screen>
1012       <para>Lists the pools defined for the mounted Lustre file system
1013       <literal>/mnt/lustre</literal></para>
1014       <screen>
1015 $ lfs poollist /mnt/lustre/
1016 </screen>
1017       <para>Lists the OSTs which are members of the pool
1018       <literal>my_pool</literal> in file system
1019       <literal>my_fs</literal></para>
1020       <screen>
1021 $ lfs poollist my_fs.my_pool
1022 </screen>
1023       <para>Finds all directories/files associated with
1024       <literal>poolA</literal>.</para>
1025       <screen>
1026 $ lfs find /mnt/lustre --pool poolA
1027 </screen>
1028       <para>Finds all directories/files not associated with a pool.</para>
1029       <screen>
1030 $ lfs find /mnt//lustre --pool ""
1031 </screen>
1032       <para>Finds all directories/files associated with pool.</para>
1033       <screen>
1034 $ lfs find /mnt/lustre ! --pool ""
1035 </screen>
1036       <para>Associates a directory with the pool
1037       <literal>my_pool</literal>, so all new files and directories are created
1038       in the pool.</para>
1039       <screen>
1040 $ lfs setstripe --pool my_pool /mnt/lustre/dir
1041 </screen>
1042     </section>
1043     <section remap="h5">
1044       <title>See Also</title>
1045       <para>
1046         <xref linkend="dbdoclet.50438219_38274" />
1047       </para>
1048     </section>
1049   </section>
1050   <section xml:id="dbdoclet.50438206_42260">
1051     <title>
1052       <indexterm>
1053         <primary>lfs_migrate</primary>
1054       </indexterm>
1055       <literal>lfs_migrate</literal>
1056     </title>
1057     <para>The
1058     <literal>lfs_migrate</literal> utility is a simple tool to migrate files
1059     between Lustre OSTs.</para>
1060     <section remap="h5">
1061       <title>Synopsis</title>
1062       <screen>
1063 lfs_migrate [-c <replaceable>stripecount</replaceable>] [-h] [-l] [-n] [-q] [-R] [-s] [-y]
1064 [file|directory ...]
1065 </screen>
1066     </section>
1067     <section remap="h5">
1068       <title>Description</title>
1069       <para>The
1070       <literal>lfs_migrate</literal> utility is a simple tool to assist
1071       migration of files between Lustre OSTs. The utility copies each specified
1072       file to a new file, verifies the file contents have not changed, and then
1073       renames the new file to the original filename. This allows balanced space
1074       usage between OSTs, moving files off OSTs that are starting to show
1075       hardware problems (though are still functional) or OSTs that will be
1076       discontinued.</para>
1077       <warning>
1078         <para>For versions of Lustre before 2.5,
1079         <literal>lfs_migrate</literal> is not closely integrated with the MDS,
1080         it cannot determine whether a file is currently open and/or in-use by
1081         other applications or nodes. This makes it UNSAFE for use on files that
1082         might be modified by other applications, since the migrated file is
1083         only a copy of the current file. This results in the old file becoming
1084         an open-unlinked file and any modifications to that file are
1085         lost.</para>
1086       </warning>
1087       <para>Files to be migrated can be specified as command-line arguments. If
1088       a directory is specified on the command-line then all files within the
1089       directory are migrated. If no files are specified on the command-line,
1090       then a list of files is read from the standard input, making
1091       <literal>lfs_migrate</literal> suitable for use with
1092       <literal>lfs</literal> find to locate files on specific OSTs and/or
1093       matching other file attributes.</para>
1094       <para>The current file allocation policies on the MDS dictate where the
1095       new files are placed, taking into account whether specific OSTs have been
1096       disabled on the MDS via
1097       <literal>lctl</literal>(preventing new files from being allocated there),
1098       whether some OSTs are overly full (reducing the number of files placed on
1099       those OSTs), or if there is a specific default file striping for the
1100       target directory (potentially changing the stripe count, stripe size, OST
1101       pool, or OST index of a new file).</para>
1102       <note>
1103         <para>The
1104         <literal>lfs_migrate</literal> utility can also be used in some cases to
1105         reduce file
1106         <indexterm>
1107           <primary>fragmentation</primary>
1108         </indexterm>fragmentation. File fragmentation will typically reduce
1109         Lustre file system performance. File fragmentation may be observed on
1110         an aged file system and will commonly occur if the file was written by
1111         many threads. Provided there is sufficient free space (or if it was
1112         written when the file system was nearly full) that is less fragmented
1113         than the file being copied, re-writing a file with
1114         <literal>lfs_migrate</literal> will result in a migrated file with
1115         reduced fragmentation. The tool
1116         <literal>filefrag</literal> can be used to report file fragmentation.
1117         See
1118         <xref linkend="dbdoclet.50438206_75125" /></para>
1119       </note>
1120       <note>
1121         <para>As long as a file has extent lengths of tens of megabytes (
1122         <replaceable>read_bandwidth * seek_time</replaceable>) or more, the
1123         read performance for the file will not be significantly impacted by
1124         fragmentation, since the read pipeline can be filled by large reads
1125         from disk even with an occasional disk seek.</para>
1126       </note>
1127     </section>
1128     <section remap="h5">
1129       <title>Options</title>
1130       <para>Options supporting
1131       <literal>lfs_migrate</literal> are described below.</para>
1132       <informaltable frame="all">
1133         <tgroup cols="2">
1134           <colspec colname="c1" colwidth="50*" />
1135           <colspec colname="c2" colwidth="50*" />
1136           <thead>
1137             <row>
1138               <entry>
1139                 <para>
1140                   <emphasis role="bold">Option</emphasis>
1141                 </para>
1142               </entry>
1143               <entry>
1144                 <para>
1145                   <emphasis role="bold">Description</emphasis>
1146                 </para>
1147               </entry>
1148             </row>
1149           </thead>
1150           <tbody>
1151             <row>
1152               <entry>
1153                 <para>
1154                   <literal>-c
1155                   <replaceable>stripecount</replaceable></literal>
1156                 </para>
1157               </entry>
1158               <entry>
1159                 <para>Restripe file using the specified stripe count. This
1160                 option may not be specified at the same time as the
1161                 <literal>-R</literal> option.</para>
1162               </entry>
1163             </row>
1164             <row>
1165               <entry>
1166                 <para>
1167                   <literal>-h</literal>
1168                 </para>
1169               </entry>
1170               <entry>
1171                 <para>Display help information.</para>
1172               </entry>
1173             </row>
1174             <row>
1175               <entry>
1176                 <literal>-l</literal>
1177               </entry>
1178               <entry>
1179                 <para>Migrate files with hard links (skips, by default). Files
1180                 with multiple hard links are split into multiple separate files
1181                 by
1182                 <literal>lfs_migrate</literal>, so they are skipped, by
1183                 default, to avoid breaking the hard links.</para>
1184               </entry>
1185             </row>
1186             <row>
1187               <entry>
1188                 <literal>-n</literal>
1189               </entry>
1190               <entry>
1191                 <para>Only print the names of files to be migrated.</para>
1192               </entry>
1193             </row>
1194             <row>
1195               <entry>
1196                 <para>
1197                   <literal>-q</literal>
1198                 </para>
1199               </entry>
1200               <entry>
1201                 <para>Run quietly (does not print filenames or status).</para>
1202               </entry>
1203             </row>
1204             <row>
1205               <entry>
1206                 <literal>-R</literal>
1207               </entry>
1208               <entry>Restripe file using default directory striping instead of
1209               keeping striping. This option may not be specified at the same
1210               time as the
1211               <literal>-c</literal> option.</entry>
1212             </row>
1213             <row>
1214               <entry>
1215                 <literal>-s</literal>
1216               </entry>
1217               <entry>Skip file data comparison after migrate. Default is to
1218               compare migrated file against original to verify
1219               correctness.</entry>
1220             </row>
1221             <row>
1222               <entry>
1223                 <para>
1224                   <literal>-y</literal>
1225                 </para>
1226               </entry>
1227               <entry>
1228                 <para>Answer '
1229                 <literal>y</literal>' to usage warning without prompting (for
1230                 scripts, use with caution).</para>
1231               </entry>
1232             </row>
1233           </tbody>
1234         </tgroup>
1235       </informaltable>
1236     </section>
1237     <section remap="h5">
1238       <title>Examples</title>
1239       <para>Rebalances all files in
1240       <literal>/mnt/lustre/dir</literal>.</para>
1241       <screen>
1242 $ lfs_migrate /mnt/lustre/file
1243 </screen>
1244       <para>Migrates files in /test filesystem on OST004 larger than 4 GB in
1245       size.</para>
1246       <screen>
1247 $ lfs find /test -obd test-OST004 -size +4G | lfs_migrate -y
1248 </screen>
1249     </section>
1250     <section remap="h5">
1251       <title>See Also</title>
1252       <para>
1253         <xref linkend="dbdoclet.50438206_94597" />
1254       </para>
1255     </section>
1256   </section>
1257   <section xml:id="dbdoclet.50438206_75125">
1258     <title>
1259       <indexterm>
1260         <primary>filefrag</primary>
1261       </indexterm>
1262       <literal>filefrag</literal>
1263     </title>
1264     <para>The
1265     <literal>e2fsprogs</literal> package contains the
1266     <literal>filefrag</literal> tool which reports the extent of file
1267     fragmentation.</para>
1268     <section remap="h5">
1269       <title>Synopsis</title>
1270       <screen>
1271 filefrag [ -belsv ] [ files...  ]
1272 </screen>
1273     </section>
1274     <section remap="h5">
1275       <title>Description</title>
1276       <para>The
1277       <literal>filefrag</literal> utility reports the extent of fragmentation in
1278       a given file. The
1279       <literal>filefrag</literal> utility obtains the extent information from
1280       Lustre files using the
1281       <literal>FIEMAP ioctl</literal>, which is efficient and fast, even for
1282       very large files.</para>
1283       <para>In default mode
1284       <footnote>
1285         <para>The default mode is faster than the verbose/extent mode since it
1286         only counts the number of extents.</para>
1287       </footnote>,
1288       <literal>filefrag</literal> prints the number of physically discontiguous
1289       extents in the file. In extent or verbose mode, each extent is printed
1290       with details such as the blocks allocated on each OST. For a Lustre file
1291       system, the extents are printed in device offset order (i.e. all of the
1292       extents for one OST first, then the next OST, etc.), not file logical
1293       offset order. If the file logical offset order was used, the Lustre
1294       striping would make the output very verbose and difficult to see if there
1295       was file fragmentation or not.</para>
1296       <note>
1297         <para>Note that as long as a file has extent lengths of tens of
1298         megabytes or more (i.e.
1299         <replaceable>read_bandwidth * seek_time &gt;
1300         extent_length</replaceable>), the read performance for the file will
1301         not be significantly impacted by fragmentation, since file readahead
1302         can fully utilize the disk disk bandwidth even with occasional
1303         seeks.</para>
1304       </note>
1305       <para>In default mode
1306       <footnote>
1307         <para>The default mode is faster than the verbose/extent mode.</para>
1308       </footnote>,
1309       <literal>filefrag</literal> returns the number of physically discontiguous
1310       extents in the file. In extent or verbose mode, each extent is printed
1311       with details. For a Lustre file system, the extents are printed in device
1312       offset order, not logical offset order.</para>
1313     </section>
1314     <section remap="h5">
1315       <title>Options</title>
1316       <para>The options and descriptions for the
1317       <literal>filefrag</literal> utility are listed below.</para>
1318       <informaltable frame="all">
1319         <tgroup cols="2">
1320           <colspec colname="c1" colwidth="50*" />
1321           <colspec colname="c2" colwidth="50*" />
1322           <thead>
1323             <row>
1324               <entry>
1325                 <para>
1326                   <emphasis role="bold">Option</emphasis>
1327                 </para>
1328               </entry>
1329               <entry>
1330                 <para>
1331                   <emphasis role="bold">Description</emphasis>
1332                 </para>
1333               </entry>
1334             </row>
1335           </thead>
1336           <tbody>
1337             <row>
1338               <entry>
1339                 <para>
1340                   <literal>-b</literal>
1341                 </para>
1342               </entry>
1343               <entry>
1344                 <para>Uses the 1024-byte blocksize for the output. By default,
1345                 this blocksize is used by the Lustre file system, since OSTs
1346                 may use different block sizes.</para>
1347               </entry>
1348             </row>
1349             <row>
1350               <entry>
1351                 <para>
1352                   <literal>-e</literal>
1353                 </para>
1354               </entry>
1355               <entry>
1356                 <para>Uses the extent mode when printing the output. This is
1357                 the default for Lustre files in verbose mode.</para>
1358               </entry>
1359             </row>
1360             <row>
1361               <entry>
1362                 <para>
1363                   <literal>-l</literal>
1364                 </para>
1365               </entry>
1366               <entry>
1367                 <para>Displays extents in LUN offset order. This is the only
1368                 available mode for Lustre.</para>
1369               </entry>
1370             </row>
1371             <row>
1372               <entry>
1373                 <para>
1374                   <literal>-s</literal>
1375                 </para>
1376               </entry>
1377               <entry>
1378                 <para>Synchronizes any unwritten file data to disk before
1379                 requesting the mapping.</para>
1380               </entry>
1381             </row>
1382             <row>
1383               <entry>
1384                 <para>
1385                   <literal>-v</literal>
1386                 </para>
1387               </entry>
1388               <entry>
1389                 <para>Prints the file's layout in verbose mode when checking
1390                 file fragmentation, including the logical to physical mapping
1391                 for each extent in the file and the OST index.</para>
1392               </entry>
1393             </row>
1394           </tbody>
1395         </tgroup>
1396       </informaltable>
1397     </section>
1398     <section remap="h5">
1399       <title>Examples</title>
1400       <para>Lists default output.</para>
1401       <screen>
1402 $ filefrag /mnt/lustre/foo
1403 /mnt/lustre/foo: 13 extents found
1404 </screen>
1405       <para>Lists verbose output in extent format.</para>
1406       <screen>
1407 $ filefrag -v /mnt/lustre/foo
1408 Filesystem type is: bd00bd0
1409 File size of /mnt/lustre/foo is 1468297786 (1433888 blocks of 1024 bytes)
1410  ext:     device_logical:        physical_offset: length:  dev: flags:
1411    0:        0..  122879: 2804679680..2804802559: 122880: 0002: network
1412    1:   122880..  245759: 2804817920..2804940799: 122880: 0002: network
1413    2:   245760..  278527: 2804948992..2804981759:  32768: 0002: network
1414    3:   278528..  360447: 2804982784..2805064703:  81920: 0002: network
1415    4:   360448..  483327: 2805080064..2805202943: 122880: 0002: network
1416    5:   483328..  606207: 2805211136..2805334015: 122880: 0002: network
1417    6:   606208..  729087: 2805342208..2805465087: 122880: 0002: network
1418    7:   729088..  851967: 2805473280..2805596159: 122880: 0002: network
1419    8:   851968..  974847: 2805604352..2805727231: 122880: 0002: network
1420    9:   974848.. 1097727: 2805735424..2805858303: 122880: 0002: network
1421   10:  1097728.. 1220607: 2805866496..2805989375: 122880: 0002: network
1422   11:  1220608.. 1343487: 2805997568..2806120447: 122880: 0002: network
1423   12:  1343488.. 1433599: 2806128640..2806218751:  90112: 0002: network
1424 /mnt/lustre/foo: 13 extents found
1425 </screen>
1426     </section>
1427   </section>
1428   <section xml:id="dbdoclet.50438206_86244">
1429     <title>
1430       <indexterm>
1431         <primary>mount</primary>
1432       </indexterm>
1433       <literal>mount</literal>
1434     </title>
1435     <para>The standard
1436     <literal>mount(8)</literal> Linux command is used to mount a Lustre file
1437     system. When mounting a Lustre file system, mount(8) executes the
1438     <literal>/sbin/mount.lustre</literal> command to complete the mount. The
1439     mount command supports these options specific to a Lustre file
1440     system:</para>
1441     <informaltable frame="all">
1442       <tgroup cols="2">
1443         <colspec colname="c1" colwidth="50*" />
1444         <colspec colname="c2" colwidth="50*" />
1445         <thead>
1446           <row>
1447             <entry>
1448               <para>
1449                 <emphasis role="bold">Server options</emphasis>
1450               </para>
1451             </entry>
1452             <entry>
1453               <para>
1454                 <emphasis role="bold">Description</emphasis>
1455               </para>
1456             </entry>
1457           </row>
1458         </thead>
1459         <tbody>
1460           <row>
1461             <entry>
1462               <para>
1463                 <literal>abort_recov</literal>
1464               </para>
1465             </entry>
1466             <entry>
1467               <para>Aborts recovery when starting a target</para>
1468             </entry>
1469           </row>
1470           <row>
1471             <entry>
1472               <para>
1473                 <literal>nosvc</literal>
1474               </para>
1475             </entry>
1476             <entry>
1477               <para>Starts only MGS/MGC servers</para>
1478             </entry>
1479           </row>
1480           <row>
1481             <entry>
1482               <para>
1483                 <literal>nomgs</literal>
1484               </para>
1485             </entry>
1486             <entry>
1487               <para>Start a MDT with a co-located MGS without starting the
1488               MGS</para>
1489             </entry>
1490           </row>
1491           <row>
1492             <entry>
1493               <para>
1494                 <literal>exclude</literal>
1495               </para>
1496             </entry>
1497             <entry>
1498               <para>Starts with a dead OST</para>
1499             </entry>
1500           </row>
1501           <row>
1502             <entry>
1503               <para>
1504                 <literal>md_stripe_cache_size</literal>
1505               </para>
1506             </entry>
1507             <entry>
1508               <para>Sets the stripe cache size for server side disk with a
1509               striped raid configuration</para>
1510             </entry>
1511           </row>
1512         </tbody>
1513       </tgroup>
1514     </informaltable>
1515     <informaltable frame="all">
1516       <tgroup cols="2">
1517         <colspec colname="c1" colwidth="50*" />
1518         <colspec colname="c2" colwidth="50*" />
1519         <thead>
1520           <row>
1521             <entry>
1522               <para>
1523                 <emphasis role="bold">Client options</emphasis>
1524               </para>
1525             </entry>
1526             <entry>
1527               <para>
1528                 <emphasis role="bold">Description</emphasis>
1529               </para>
1530             </entry>
1531           </row>
1532         </thead>
1533         <tbody>
1534           <row>
1535             <entry>
1536               <para>
1537                 <literal>flock/noflock/localflock</literal>
1538               </para>
1539             </entry>
1540             <entry>
1541               <para>Enables/disables global flock or local flock support</para>
1542             </entry>
1543           </row>
1544           <row>
1545             <entry>
1546               <para>
1547                 <literal>user_xattr/nouser_xattr</literal>
1548               </para>
1549             </entry>
1550             <entry>
1551               <para>Enables/disables user-extended attributes</para>
1552             </entry>
1553           </row>
1554           <row>
1555             <entry>
1556               <para>
1557                 <literal>user_fid2path/nouser_fid2path</literal>
1558               </para>
1559             </entry>
1560             <entry>
1561               <para condition="l23">Enables/disables FID to path translation by
1562               regular users</para>
1563             </entry>
1564           </row>
1565           <row>
1566             <entry>
1567               <para>
1568                 <literal>retry=</literal>
1569               </para>
1570             </entry>
1571             <entry>
1572               <para>Number of times a client will retry to mount the file
1573               system</para>
1574             </entry>
1575           </row>
1576         </tbody>
1577       </tgroup>
1578     </informaltable>
1579   </section>
1580   <section xml:id="dbdoclet.50438206_56217">
1581     <title>Handling Timeouts</title>
1582     <para>Timeouts are the most common cause of hung applications. After a
1583     timeout involving an MDS or failover OST, applications attempting to access
1584     the disconnected resource wait until the connection gets
1585     established.</para>
1586     <para>When a client performs any remote operation, it gives the server a
1587     reasonable amount of time to respond. If a server does not reply either due
1588     to a down network, hung server, or any other reason, a timeout occurs which
1589     requires a recovery.</para>
1590     <para>If a timeout occurs, a message (similar to this one), appears on the
1591     console of the client, and in
1592     <literal>/var/log/messages</literal>:</para>
1593     <screen>
1594 LustreError: 26597:(client.c:810:ptlrpc_expire_one_request()) @@@ timeout
1595
1596 req@a2d45200 x5886/t0 o38-&gt;mds_svc_UUID@NID_mds_UUID:12 lens 168/64 ref 1 fl
1597
1598 RPC:/0/0 rc 0
1599 </screen>
1600   </section>
1601 </chapter>