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 xmlns:xi="http://www.w3.org/2001/XInclude">
   6   <title xml:id="userutilities.title">User Utilities</title>
   7   <para>This chapter describes user utilities.</para>
   8   <section xml:id="dbdoclet.50438206_94597">
   9     <title>
  10       <indexterm>
  11         <primary>lfs</primary>
  12       </indexterm>
  13       <literal>lfs</literal>
  14     </title>
  15     <para>The
  16     <literal>lfs</literal> utility can be used for user configuration routines
  17     and monitoring.</para>
  18     <section remap="h5">
  19       <title>Synopsis</title>
  20       <screen>
  21 lfs
  22 lfs changelog [--follow] <replaceable>mdt_name</replaceable> [startrec [endrec]]
  23 lfs changelog_clear <replaceable>mdt_name id endrec</replaceable>
  24 lfs check <replaceable>mds|osts|servers</replaceable>
  25 lfs df [-i] [-h] [--pool]-p <replaceable>fsname</replaceable>[.<replaceable>pool</replaceable>] [<replaceable>path</replaceable>] [--lazy]
  26 lfs find [[!] --atime|-A [-+]N] [[!] --mtime|-M [-+]N]
  27          [[!] --ctime|-C [-+]N] [--maxdepth|-D N] [--name|-n <replaceable>pattern</replaceable>]
  28          [--print|-p] [--print0|-P] [[!] --obd|-O <replaceable>ost_name</replaceable>[,<replaceable>ost_name...</replaceable>]]
  29          [[!] --size|-S [+-]N[kMGTPE]] --type |-t {bcdflpsD}]
  30          [[!] --gid|-g|--group|-G <replaceable>gname|gid</replaceable>]
  31          [[!] --uid|-u|--user|-U <replaceable>uname|uid</replaceable>]
  32          <replaceable>dirname|filename</replaceable>
  33 lfs getname [-h]|[path...]
  34 lfs getstripe [--obd|-O <replaceable>ost_name</replaceable>] [--quiet|-q] [--verbose|-v]
  35               [--stripe-count|-c] [--stripe-index|-i]
  36               [--stripe-size|-s] [--pool|-p] [--directory|-d]
  37               [--mdt-index|-M] [--recursive|-r] [--raw|-R]
  38               [--layout|-L]
  39               <replaceable>dirname|filename</replaceable> ...
  40 lfs setstripe [--size|-s stripe_size] [--count|-c <replaceable>stripe_count</replaceable>]
  41               [--stripe-index|-i <replaceable>start_ost_index</replaceable>]
  42               [--ost-list|-o <replaceable>ost_indicies</replaceable>]
  43               [--pool|-p <replaceable>pool</replaceable>]
  44               <replaceable>dirname|filename</replaceable>
  45 lfs setstripe -d <replaceable>dir</replaceable>
  46 lfs osts [path]
  47 lfs pool_list <replaceable>filesystem</replaceable>[.<replaceable>pool</replaceable>]| <replaceable>pathname</replaceable>
  48 lfs quota [-q] [-v] [-h] [-o <replaceable>obd_uuid</replaceable>|-I <replaceable>ost_idx</replaceable>|-i <replaceable>mdt_idx</replaceable>]
  49           [-u <replaceable>username|uid|-g</replaceable> <replaceable>group|gid</replaceable>] <replaceable>/mount_point</replaceable>
  50 lfs quota -t -u|-g <replaceable>/mount_point</replaceable>
  51 lfs quotacheck [-ug] <replaceable>/mount_point</replaceable>
  52 lfs quotachown [-i] <replaceable>/mount_point</replaceable>
  53 lfs quotainv [-ug] [-f] <replaceable>/mount_point</replaceable>
  54 lfs quotaon [-ugf] <replaceable>/mount_point</replaceable>
  55 lfs quotaoff [-ug] <replaceable>/mount_point</replaceable>
  56 lfs setquota {-u|--user|-g|--group} <replaceable>uname|uid|gname|gid</replaceable>
  57              [--block-softlimit <replaceable>block_softlimit</replaceable>]
  58              [--block-hardlimit <replaceable>block_hardlimit</replaceable>]
  59              [--inode-softlimit <replaceable>inode_softlimit</replaceable>]
  60              [--inode-hardlimit <replaceable>inode_hardlimit</replaceable>]
  61              <replaceable>/mount_point</replaceable>
  62 lfs setquota -u|--user|-g|--group <replaceable>uname|uid|gname|gid</replaceable>
  63              [-b <replaceable>block_softlimit</replaceable>] [-B <replaceable>block_hardlimit</replaceable>]
  64              [-i <replaceable>inode-softlimit</replaceable>] [-I <replaceable>inode_hardlimit</replaceable>]
  65              <replaceable>/mount_point</replaceable>
  66 lfs setquota -t -u|-g [--block-grace <replaceable>block_grace</replaceable>]
  67              [--inode-grace <replaceable>inode_grace</replaceable>]
  68              <replaceable>/mount_point</replaceable>
  69 lfs setquota -t -u|-g [-b <replaceable>block_grace</replaceable>] [-i <replaceable>inode_grace</replaceable>]
  70              <replaceable>/mount_point</replaceable>
  71 lfs help
  72 </screen>
  73       <note>
  74         <para>In the above example, the
  75         <literal>
  76           <replaceable>/mount_point</replaceable>
  77         </literal> parameter refers to the mount point of the Lustre file
  78         system.</para>
  79       </note>
  80       <note>
  81         <para>The old lfs quota output was very detailed and contained
  82         cluster-wide quota statistics (including cluster-wide limits for a
  83         user/group and cluster-wide usage for a user/group), as well as
  84         statistics for each MDS/OST. Now,
  85         <literal>lfs quota</literal> has been updated to provide only
  86         cluster-wide statistics, by default. To obtain the full report of
  87         cluster-wide limits, usage and statistics, use the
  88         <literal>-v</literal> option with
  89         <literal>lfs quota</literal>.</para>
  90       </note>
  91     </section>
  92     <section remap="h5">
  93       <title>Description</title>
  94       <para>The
  95       <literal>lfs</literal> utility is used to create a new file with a
  96       specific striping pattern, determine the default striping pattern, gather
  97       the extended attributes (object numbers and location) for a specific
  98       file, find files with specific attributes, list OST information or set
  99       quota limits. It can be invoked interactively without any arguments or in
 100       a non-interactive mode with one of the supported arguments.</para>
 101     </section>
 102     <section remap="h5">
 103       <title>Options</title>
 104       <para>The various
 105       <literal>lfs</literal> options are listed and described below. For a
 106       complete list of available options, type help at the
 107       <literal>lfs</literal> prompt.</para>
 108       <informaltable frame="all">
 109         <tgroup cols="3">
 110           <colspec colname="c1" colwidth="20*" />
 111           <colspec colname="c2" colwidth="30*" />
 112           <colspec colname="c3" colwidth="50*" />
 113           <thead>
 114             <row>
 115               <entry nameend="c2" namest="c1">
 116                 <para>
 117                   <emphasis role="bold">Option</emphasis>
 118                 </para>
 119               </entry>
 120               <entry>
 121                 <para>
 122                   <emphasis role="bold">Description</emphasis>
 123                 </para>
 124               </entry>
 125             </row>
 126           </thead>
 127           <tbody>
 128             <row>
 129               <entry nameend="c2" namest="c1">
 130                 <para>
 131                   <literal>changelog</literal>
 132                 </para>
 133               </entry>
 134               <entry>
 135                 <para>Shows the metadata changes on an MDT. Start and end
 136                 points are optional. The
 137                 <literal>--follow</literal> option blocks on new changes; this
 138                 option is only valid when run directly on the MDT node.</para>
 139               </entry>
 140             </row>
 141             <row>
 142               <entry nameend="c2" namest="c1">
 143                 <para>
 144                   <literal>changelog_clear</literal>
 145                 </para>
 146               </entry>
 147               <entry>
 148                 <para>Indicates that changelog records previous to
 149                 <literal>
 150                   <replaceable>endrec</replaceable>
 151                 </literal> are no longer of interest to a particular consumer
 152                 <literal>
 153                   <replaceable>id</replaceable>
 154                 </literal>, potentially allowing the MDT to free up disk space.
 155                 An
 156                 <literal>
 157                   <replaceable>endrec</replaceable>
 158                 </literal> of 0 indicates the current last record. Changelog
 159                 consumers must be registered on the MDT node using
 160                 <literal>lctl</literal>.</para>
 161               </entry>
 162             </row>
 163             <row>
 164               <entry nameend="c2" namest="c1">
 165                 <literal>check</literal>
 166               </entry>
 167               <entry>
 168                 <para>Displays the status of MDS or OSTs (as specified in the
 169                 command) or all servers (MDS and OSTs).</para>
 170               </entry>
 171             </row>
 172             <row>
 173               <entry nameend="c2" namest="c1">
 174                 <literal>df [-i] [-h] [--pool|-p
 175                 <replaceable>fsname</replaceable>[.
 176                 <replaceable>pool</replaceable>] [
 177                 <replaceable>path</replaceable>] [--lazy]</literal>
 178               </entry>
 179               <entry>
 180                 <para>Use
 181                 <literal>-i</literal> to report file system disk space usage or
 182                 inode usage of each MDT or OST or, if a pool is specified with
 183                 the
 184                 <literal>-p</literal> option, a subset of OSTs.</para>
 185                 <para>By default, the usage of all mounted Lustre file systems
 186                 is reported. If the
 187                 <literal>path</literal> option is included, only the usage for
 188                 the specified file system is reported. If the
 189                 <literal>-h</literal> option is included, the output is printed
 190                 in human-readable format, using SI base-2 suffixes for
 191                 <emphasis role="bold">M</emphasis>ega-,
 192                 <emphasis role="bold">G</emphasis>iga-,
 193                 <emphasis role="bold">T</emphasis>era-,
 194                 <emphasis role="bold">P</emphasis>eta-, or
 195                 <emphasis role="bold">E</emphasis>xabytes.</para>
 196                 <para>If the
 197                 <literal>--lazy</literal> option is specified, any OST that is
 198                 currently disconnected from the client will be skipped. Using
 199                 the
 200                 <literal>--lazy</literal> option prevents the
 201                 <literal>df</literal> output from being blocked when an OST is
 202                 offline. Only the space on the OSTs that can currently be
 203                 accessed are returned. The
 204                 <literal>llite.*.lazystatfs</literal> tunable can be enabled to
 205                 make this the default behaviour for all
 206                 <literal>statfs()</literal> operations.</para>
 207               </entry>
 208             </row>
 209             <row>
 210               <entry nameend="c2" namest="c1">
 211                 <para>
 212                   <literal>find</literal>
 213                 </para>
 214               </entry>
 215               <entry>
 216                 <para>Searches the directory tree rooted at the given
 217                 directory/filename for files that match the given
 218                 parameters.</para>
 219                 <para>Using
 220                 <literal>!</literal> before an option negates its meaning (files
 221                 NOT matching the parameter). Using
 222                 <literal>+</literal> before a numeric value means files with the
 223                 parameter OR MORE. Using
 224                 <literal>-</literal> before a numeric value means files with the
 225                 parameter OR LESS.</para>
 226               </entry>
 227             </row>
 228             <row>
 229               <entry />
 230               <entry>
 231                 <literal>--atime</literal>
 232               </entry>
 233               <entry>
 234                 <para>File was last accessed N*24 hours ago. (There is no
 235                 guarantee that
 236                 <literal>atime</literal> is kept coherent across the
 237                 cluster.)</para>
 238                 <para>OSTs store a transient
 239                 <literal>atime</literal> that is updated when clients do read
 240                 requests. Permanent
 241                 <literal>atime</literal> is written to the MDS when the file is
 242                 closed. However, on-disk atime is only updated if it is more
 243                 than 60 seconds old (
 244                 <literal>/proc/fs/lustre/mds/*/max_atime_diff</literal>). The
 245                 Lustre software considers the latest
 246                 <literal>atime</literal> from all OSTs. If a
 247                 <literal>setattr</literal> is set by user, then it is updated on
 248                 both the MDS and OST, allowing the
 249                 <literal>atime</literal> to go backward.</para>
 250               </entry>
 251             </row>
 252             <row>
 253               <entry>
 254                 <para>&#160;</para>
 255               </entry>
 256               <entry>
 257                 <para>
 258                   <literal>--ctime</literal>
 259                 </para>
 260               </entry>
 261               <entry>
 262                 <para>File status was last changed N*24 hours ago.</para>
 263               </entry>
 264             </row>
 265             <row>
 266               <entry>
 267                 <para>&#160;</para>
 268               </entry>
 269               <entry>
 270                 <para>
 271                   <literal>--mtime</literal>
 272                 </para>
 273               </entry>
 274               <entry>
 275                 <para>File data was last modified N*24 hours ago.</para>
 276               </entry>
 277             </row>
 278             <row>
 279               <entry>
 280                 <para>&#160;</para>
 281               </entry>
 282               <entry>
 283                 <para>
 284                   <literal>--obd</literal>
 285                 </para>
 286               </entry>
 287               <entry>
 288                 <para>File has an object on a specific OST(s).</para>
 289               </entry>
 290             </row>
 291             <row>
 292               <entry>
 293                 <para>&#160;</para>
 294               </entry>
 295               <entry>
 296                 <para>
 297                   <literal>--size</literal>
 298                 </para>
 299               </entry>
 300               <entry>
 301                 <para>File has a size in bytes, or kilo-, Mega-, Giga-, Tera-,
 302                 Peta- or Exabytes if a suffix is given.</para>
 303               </entry>
 304             </row>
 305             <row>
 306               <entry>
 307                 <para>&#160;</para>
 308               </entry>
 309               <entry>
 310                 <para>
 311                   <literal>--type</literal>
 312                 </para>
 313               </entry>
 314               <entry>
 315                 <para>File has the type - block, character, directory, pipe,
 316                 file, symlink, socket or door (used in Solaris operating
 317                 system).</para>
 318               </entry>
 319             </row>
 320             <row>
 321               <entry>
 322                 <para>&#160;</para>
 323               </entry>
 324               <entry>
 325                 <para>
 326                   <literal>--uid</literal>
 327                 </para>
 328               </entry>
 329               <entry>
 330                 <para>File has a specific numeric user ID.</para>
 331               </entry>
 332             </row>
 333             <row>
 334               <entry>
 335                 <para>&#160;</para>
 336               </entry>
 337               <entry>
 338                 <para>
 339                   <literal>--user</literal>
 340                 </para>
 341               </entry>
 342               <entry>
 343                 <para>File owned by a specific user (numeric user ID
 344                 allowed).</para>
 345               </entry>
 346             </row>
 347             <row>
 348               <entry>
 349                 <para>&#160;</para>
 350               </entry>
 351               <entry>
 352                 <para>
 353                   <literal>--gid</literal>
 354                 </para>
 355               </entry>
 356               <entry>
 357                 <para>File has a specific group ID.</para>
 358               </entry>
 359             </row>
 360             <row>
 361               <entry>
 362                 <para>&#160;</para>
 363               </entry>
 364               <entry>
 365                 <para>
 366                   <literal>--group</literal>
 367                 </para>
 368               </entry>
 369               <entry>
 370                 <para>File belongs to a specific group (numeric group ID
 371                 allowed).</para>
 372               </entry>
 373             </row>
 374             <row>
 375               <entry>
 376                 <para>&#160;</para>
 377               </entry>
 378               <entry>
 379                 <para>-
 380                 <literal>-maxdepth</literal></para>
 381               </entry>
 382               <entry>
 383                 <para>Limits find to descend at most N levels of the directory
 384                 tree.</para>
 385               </entry>
 386             </row>
 387             <row>
 388               <entry>
 389                 <para>&#160;</para>
 390               </entry>
 391               <entry>
 392                 <para>
 393                 <literal>--print</literal>/
 394                 <literal>--print0</literal></para>
 395               </entry>
 396               <entry>
 397                 <para>Prints the full filename, followed by a new line or NULL
 398                 character correspondingly.</para>
 399               </entry>
 400             </row>
 401             <row>
 402               <entry nameend="c2" namest="c1">
 403                 <para>
 404                   <literal>osts [path]</literal>
 405                 </para>
 406               </entry>
 407               <entry>
 408                 <para>Lists all OSTs for the file system. If a path located on
 409                 a mounted Lustre file system is specified, then only OSTs
 410                 belonging to this file system are displayed.</para>
 411               </entry>
 412             </row>
 413             <row>
 414               <entry nameend="c2" namest="c1">
 415                 <para>
 416                   <literal>getname [path...]</literal>
 417                 </para>
 418               </entry>
 419               <entry>
 420                 <para>List each Lustre file system instance associated with
 421                 each Lustre mount point. If no path is specified, all Lustre
 422                 mount points are interrogated. If a list of paths is provided,
 423                 the instance of each path is provided. If the path is not a
 424                 Lustre instance 'No such device' is returned.</para>
 425               </entry>
 426             </row>
 427             <row>
 428               <entry nameend="c2" namest="c1">
 429                 <para>
 430                   <literal>getstripe</literal>
 431                 </para>
 432               </entry>
 433               <entry>
 434                 <para>Lists striping information for a given filename or
 435                 directory. By default, the stripe count, stripe size and offset
 436                 are returned.</para>
 437                 <para>If you only want specific striping information, then the
 438                 options of
 439                 <literal>--stripe-count</literal>,
 440                 <literal>--stripe-size</literal>,
 441                 <literal>--stripe-index</literal>,
 442                 <literal>--layout</literal>, or
 443                 <literal>--pool</literal> plus various combinations of these
 444                 options can be used to retrieve specific information.</para>
 445                 <para>If the
 446                 <literal>--raw</literal> option is specified, the stripe
 447                 information is printed without substituting the file system
 448                 default values for unspecified fields. If the striping EA is
 449                 not set, 0, 0, and -1 will be printed for the stripe count,
 450                 size, and offset respectively.</para>
 451                 <para condition="l24">The
 452                 <literal>--mdt-index</literal> prints the index of the MDT for a given
 453                 directory. See
 454                 <xref linkend="dbdoclet.rmremotedir" />.</para>
 455               </entry>
 456             </row>
 457             <row>
 458               <entry>
 459                 <para>&#160;</para>
 460               </entry>
 461               <entry>
 462                 <para>
 463                   <literal>--obd
 464                   <replaceable>ost_name</replaceable></literal>
 465                 </para>
 466               </entry>
 467               <entry>
 468                 <para>Lists files that have an object on a specific OST.</para>
 469               </entry>
 470             </row>
 471             <row>
 472               <entry>
 473                 <para>&#160;</para>
 474               </entry>
 475               <entry>
 476                 <para>
 477                   <literal>--quiet</literal>
 478                 </para>
 479               </entry>
 480               <entry>
 481                 <para>Lists details about the file's object ID
 482                 information.</para>
 483               </entry>
 484             </row>
 485             <row>
 486               <entry>
 487                 <para>&#160;</para>
 488               </entry>
 489               <entry>
 490                 <para>
 491                   <literal>--verbose</literal>
 492                 </para>
 493               </entry>
 494               <entry>
 495                 <para>Prints additional striping information.</para>
 496               </entry>
 497             </row>
 498             <row>
 499               <entry>
 500                 <para>&#160;</para>
 501               </entry>
 502               <entry>
 503                 <para>
 504                   <literal>--count</literal>
 505                 </para>
 506               </entry>
 507               <entry>
 508                 <para>Lists the stripe count (how many OSTs to use).</para>
 509               </entry>
 510             </row>
 511             <row>
 512               <entry>
 513                 <para>&#160;</para>
 514               </entry>
 515               <entry>
 516                 <para>
 517                   <literal>--index</literal>
 518                 </para>
 519               </entry>
 520               <entry>
 521                 <para>Lists the index for each OST in the file system.</para>
 522               </entry>
 523             </row>
 524             <row>
 525               <entry>
 526                 <para>&#160;</para>
 527               </entry>
 528               <entry>
 529                 <para>
 530                   <literal>--offset</literal>
 531                 </para>
 532               </entry>
 533               <entry>
 534                 <para>Lists the OST index on which file striping starts.</para>
 535               </entry>
 536             </row>
 537             <row>
 538               <entry>
 539                 <para>&#160;</para>
 540               </entry>
 541               <entry>
 542                 <para>
 543                   <literal>--pool</literal>
 544                 </para>
 545               </entry>
 546               <entry>
 547                 <para>Lists the pools to which a file belongs.</para>
 548               </entry>
 549             </row>
 550             <row>
 551               <entry>
 552                 <para>&#160;</para>
 553               </entry>
 554               <entry>
 555                 <para>
 556                   <literal>--size</literal>
 557                 </para>
 558               </entry>
 559               <entry>
 560                 <para>Lists the stripe size (how much data to write to one OST
 561                 before moving to the next OST).</para>
 562               </entry>
 563             </row>
 564             <row>
 565               <entry>
 566                 <para>&#160;</para>
 567               </entry>
 568               <entry>
 569                 <para>
 570                   <literal>--directory</literal>
 571                 </para>
 572               </entry>
 573               <entry>
 574                 <para>Lists entries about a specified directory instead of its
 575                 contents (in the same manner as
 576                 <literal>ls -d</literal>).</para>
 577               </entry>
 578             </row>
 579             <row>
 580               <entry>
 581                 <para>&#160;</para>
 582               </entry>
 583               <entry>
 584                 <para>
 585                   <literal>--recursive</literal>
 586                 </para>
 587               </entry>
 588               <entry>
 589                 <para>Recurses into all sub-directories.</para>
 590               </entry>
 591             </row>
 592             <row>
 593               <entry nameend="c2" namest="c1">
 594                 <para>
 595                   <literal>setstripe</literal>
 596                 </para>
 597               </entry>
 598               <entry>
 599                 <para>Create new files with a specific file layout (stripe
 600                 pattern) configuration.
 601                 <footnote>
 602                   <para>The file cannot exist prior to using
 603                   <literal>setstripe</literal>. A directory must exist prior to
 604                   using
 605                   <literal>setstripe</literal>.</para>
 606                 </footnote></para>
 607               </entry>
 608             </row>
 609             <row>
 610               <entry>
 611                 <para>&#160;</para>
 612               </entry>
 613               <entry>
 614                 <para>
 615                   <literal>--count stripe_cnt</literal>
 616                 </para>
 617               </entry>
 618               <entry>
 619                 <para>Number of OSTs over which to stripe a file. A
 620                 <literal>stripe_cnt</literal> of 0 uses the file system-wide
 621                 default stripe count (default is 1). A
 622                 <literal>stripe_cnt</literal> of -1 stripes over all available
 623                 OSTs.</para>
 624               </entry>
 625             </row>
 626             <row>
 627               <entry>
 628                 <para>&#160;</para>
 629               </entry>
 630               <entry>
 631                 <para>
 632                 <literal>--size stripe_size</literal>
 633                 <footnote>
 634                   <para>The default stripe-size is 0. The default start-ost is
 635                   -1. Do NOT confuse them! If you set start-ost to 0, all new
 636                   file creations occur on OST 0 (seldom a good idea).</para>
 637                 </footnote>&#160;</para>
 638               </entry>
 639               <entry>
 640                 <para>Number of bytes to store on an OST before moving to the
 641                 next OST. A stripe_size of 0 uses the file system's default
 642                 stripe size, (default is 1 MB). Can be specified with
 643                 <emphasis role="bold">k</emphasis>(KB),
 644                 <emphasis role="bold">m</emphasis>(MB), or
 645                 <emphasis role="bold">g</emphasis>(GB), respectively.</para>
 646               </entry>
 647             </row>
 648             <row>
 649               <entry>
 650                 <para>&#160;</para>
 651               </entry>
 652               <entry>
 653                 <para>
 654                   <literal>--stripe-index start_ost_index</literal>
 655                 </para>
 656               </entry>
 657               <entry>
 658                 <para>The OST index (base 10, starting at 0) on which to start
 659                 striping for this file. A start_ost_index value of -1 allows
 660                 the MDS to choose the starting index. This is the default
 661                 value, and it means that the MDS selects the starting OST as it
 662                 wants. We strongly recommend selecting this default, as it
 663                 allows space and load balancing to be done by the MDS as
 664                 needed. The
 665                 <literal>start_ost_index</literal> value has no relevance on
 666                 whether the MDS will use round-robin or QoS weighted allocation
 667                 for the remaining stripes in the file.</para>
 668               </entry>
 669             </row>
 670             <row>
 671               <entry>
 672                 <para>&#160;</para>
 673               </entry>
 674               <entry>
 675                 <para>
 676                   <literal>--ost-index ost_indices</literal>
 677                 </para>
 678               </entry>
 679               <entry>
 680                                 <para>This option is used to specify the exact stripe
 681                 layout on the the file system. <literal>ost_indices</literal>
 682                 is a list of OSTs referenced by their indices and index ranges
 683                 separated by commas, e.g. <literal>1,2-4,7</literal>.</para>
 684               </entry>
 685             </row>
 686             <row>
 687               <entry>
 688                 <para>&#160;</para>
 689               </entry>
 690               <entry>
 691                 <para>
 692                   <literal>--pool
 693                   <replaceable>pool</replaceable></literal>
 694                 </para>
 695               </entry>
 696               <entry>
 697                 <para>Name of the pre-defined pool of OSTs (see
 698                 <xref linkend="dbdoclet.50438219_38274" />) that will be used
 699                 for striping. The
 700                 <literal>stripe_cnt</literal>,
 701                 <literal>stripe_size</literal> and
 702                 <literal>start_ost</literal> values are used as well. The
 703                 start-ost value must be part of the pool or an error is
 704                 returned.</para>
 705               </entry>
 706             </row>
 707             <row>
 708               <entry nameend="c2" namest="c1">
 709                 <para>
 710                   <literal>setstripe -d</literal>
 711                 </para>
 712               </entry>
 713               <entry>
 714                 <para>Deletes default striping on the specified
 715                 directory.</para>
 716               </entry>
 717             </row>
 718             <row>
 719               <entry nameend="c2" namest="c1">
 720                 <para>
 721                   <literal>pool_list {filesystem}[.poolname]|{pathname}</literal>
 722                 </para>
 723               </entry>
 724               <entry>
 725                 <para>Lists pools in the file system or pathname, or OSTs in
 726                 the file system's pool.</para>
 727               </entry>
 728             </row>
 729             <row>
 730               <entry nameend="c2" namest="c1">
 731                 <para>
 732                   <literal>quota [-q] [-v] [-o
 733                   <replaceable>obd_uuid</replaceable>|-i
 734                   <replaceable>mdt_idx</replaceable>|-I
 735                   <replaceable>ost_idx</replaceable>] [-u|-g
 736                   <replaceable>uname|uid|gname|gid]</replaceable>
 737                   <replaceable>/mount_point</replaceable></literal>
 738                 </para>
 739                 <para>&#160;</para>
 740               </entry>
 741               <entry>
 742                 <para>Displays disk usage and limits, either for the full file
 743                 system or for objects on a specific OBD. A user or group name
 744                 or an ID can be specified. If both user and group are omitted,
 745                 quotas for the current UID/GID are shown. The
 746                 <literal>-q</literal> option disables printing of additional
 747                 descriptions (including column titles). It fills in blank
 748                 spaces in the
 749                 <literal>grace</literal> column with zeros (when there is no
 750                 grace period set), to ensure that the number of columns is
 751                 consistent. The
 752                 <literal>-v</literal> option provides more verbose (per-OBD
 753                 statistics) output.</para>
 754               </entry>
 755             </row>
 756             <row>
 757               <entry nameend="c2" namest="c1">
 758                 <para>
 759                   <literal>quota -t
 760                   <replaceable>-u|-g</replaceable>
 761                   <replaceable>/mount_point</replaceable></literal>
 762                 </para>
 763               </entry>
 764               <entry>
 765                 <para>Displays block and inode grace times for user (
 766                 <literal>-u</literal>) or group (
 767                 <literal>-g</literal>) quotas.</para>
 768               </entry>
 769             </row>
 770             <row>
 771               <entry nameend="c2" namest="c1">
 772                 <para>
 773                   <literal>quotachown</literal>
 774                 </para>
 775               </entry>
 776               <entry>
 777                 <para>Changes the file's owner and group on OSTs of the
 778                 specified file system.</para>
 779               </entry>
 780             </row>
 781             <row>
 782               <entry nameend="c2" namest="c1">
 783                 <para>
 784                   <literal>quotacheck [-ugf]
 785                   <replaceable>/mount_point</replaceable></literal>
 786                 </para>
 787               </entry>
 788               <entry>
 789                 <para>Scans the specified file system for disk usage, and
 790                 creates or updates quota files. Options specify quota for users
 791                 (
 792                 <literal>-u</literal>), groups (
 793                 <literal>-g</literal>), and force (
 794                 <literal>-f</literal>).</para>
 795               </entry>
 796             </row>
 797             <row>
 798               <entry nameend="c2" namest="c1">
 799                 <para>
 800                   <literal>quotaon [-ugf]
 801                   <replaceable>/mount_point</replaceable></literal>
 802                 </para>
 803               </entry>
 804               <entry>
 805                 <para>Turns on file system quotas. Options specify quota for
 806                 users (
 807                 <literal>-u</literal>), groups (
 808                 <literal>-g</literal>), and force (
 809                 <literal>-f</literal>).</para>
 810               </entry>
 811             </row>
 812             <row>
 813               <entry nameend="c2" namest="c1">
 814                 <para>
 815                   <literal>quotaoff [-ugf]
 816                   <replaceable>/mount_point</replaceable></literal>
 817                 </para>
 818               </entry>
 819               <entry>
 820                 <para>Turns off file system quotas. Options specify quota for
 821                 users (
 822                 <literal>-u</literal>), groups (
 823                 <literal>-g</literal>), and force (
 824                 <literal>-f</literal>).</para>
 825               </entry>
 826             </row>
 827             <row>
 828               <entry nameend="c2" namest="c1">
 829                 <para>
 830                   <literal>quotainv [-ug] [-f]
 831                   <replaceable>/mount_point</replaceable></literal>
 832                 </para>
 833               </entry>
 834               <entry>
 835                 <para>Clears quota files (administrative quota files if used
 836                 without
 837                 <literal>-f</literal>, operational quota files otherwise), all
 838                 of their quota entries for users (
 839                 <literal>-u</literal>) or groups (
 840                 <literal>-g</literal>). After running
 841                 <literal>quotainv</literal>, you must run
 842                 <literal>quotacheck</literal> before using quotas.</para>
 843                 <caution>
 844                   <para>Use extreme caution when using this command; its
 845                   results cannot be undone.</para>
 846                 </caution>
 847               </entry>
 848             </row>
 849             <row>
 850               <entry nameend="c2" namest="c1">
 851                 <para>
 852                   <literal>setquota -u|-g
 853                   <replaceable>
 854                   uname|uid|gname|gid}</replaceable>[--block-softlimit
 855                   <replaceable>block_softlimit</replaceable>]
 856                   [--block-hardlimit
 857                   <replaceable>block_hardlimit</replaceable>]
 858                   [--inode-softlimit
 859                   <replaceable>inode_softlimit</replaceable>]
 860                   [--inode-hardlimit
 861                   <replaceable>inode_hardlimit</replaceable>]
 862                   <replaceable>/mount_point</replaceable></literal>
 863                 </para>
 864               </entry>
 865               <entry>
 866                 <para>Sets file system quotas for users or groups. Limits can
 867                 be specified with
 868                 <literal>--{block|inode}-{softlimit|hardlimit}</literal> or
 869                 their short equivalents
 870                 <literal>-b</literal>,
 871                 <literal>-B</literal>,
 872                 <literal>-i</literal>,
 873                 <literal>-I</literal>. Users can set 1, 2, 3 or 4 limits.
 874                 <footnote>
 875                   <para>The old
 876                   <literal>setquota</literal> interface is supported, but it may
 877                   be removed in a future Lustre software release.</para>
 878                 </footnote>Also, limits can be specified with special suffixes,
 879                 -b, -k, -m, -g, -t, and -p to indicate units of 1, 2^10, 2^20,
 880                 2^30, 2^40 and 2^50, respectively. By default, the block limits
 881                 unit is 1 kilobyte (1,024), and block limits are always
 882                 kilobyte-grained (even if specified in bytes). See
 883                 <xref linkend="dbdoclet.50438206_11903" />.</para>
 884               </entry>
 885             </row>
 886             <row>
 887               <entry nameend="c2" namest="c1">
 888                 <para>
 889                   <literal>setquota -t -u|-g [--block-grace
 890                   <replaceable>block_grace</replaceable>] [--inode-grace
 891                   <replaceable>inode_grace</replaceable>]
 892                   <replaceable>/mount_point</replaceable></literal>
 893                 </para>
 894               </entry>
 895               <entry>
 896                 <para>Sets the file system quota grace times for users or
 897                 groups. Grace time is specified in '
 898                 <literal>XXwXXdXXhXXmXXs</literal>' format or as an integer
 899                 seconds value. See
 900                 <xref linkend="dbdoclet.50438206_11903" />.</para>
 901               </entry>
 902             </row>
 903             <row>
 904               <entry nameend="c2" namest="c1">
 905                 <para>
 906                   <literal>help</literal>
 907                 </para>
 908               </entry>
 909               <entry>
 910                 <para>Provides brief help on various
 911                 <literal>lfs</literal> arguments.</para>
 912               </entry>
 913             </row>
 914             <row>
 915               <entry nameend="c2" namest="c1">
 916                 <para>
 917                   <literal>exit/quit</literal>
 918                 </para>
 919               </entry>
 920               <entry>
 921                 <para>Quits the interactive
 922                 <literal>lfs</literal> session.</para>
 923               </entry>
 924             </row>
 925           </tbody>
 926         </tgroup>
 927       </informaltable>
 928     </section>
 929     <section remap="h5" xml:id="dbdoclet.50438206_11903">
 930       <title>Examples</title>
 931       <para>Creates a file striped on two OSTs with 128 KB on each
 932       stripe.</para>
 933       <screen>
 934 $ lfs setstripe -s 128k -c 2 /mnt/lustre/file1
 935 </screen>
 936       <para>Deletes a default stripe pattern on a given directory. New files
 937       use the default striping pattern.</para>
 938       <screen>
 939 $ lfs setstripe -d /mnt/lustre/dir
 940 </screen>
 941       <para>Lists the detailed object allocation of a given file.</para>
 942       <screen>
 943 $ lfs getstripe -v /mnt/lustre/file1
 944 </screen>
 945       <para>List all the mounted Lustre file systems and corresponding Lustre
 946       instances.</para>
 947       <screen>
 948 $ lfs getname
 949 </screen>
 950       <para>Efficiently lists all files in a given directory and its
 951       subdirectories.</para>
 952       <screen>
 953 $ lfs find /mnt/lustre
 954 </screen>
 955       <para>Recursively lists all regular files in a given directory more than
 956       30 days old.</para>
 957       <screen>
 958 $ lfs find /mnt/lustre -mtime +30 -type f -print
 959 </screen>
 960       <para>Recursively lists all files in a given directory that have objects
 961       on OST2-UUID. The lfs check servers command checks the status of all
 962       servers (MDT and OSTs).</para>
 963       <screen>
 964 $ lfs find --obd OST2-UUID /mnt/lustre/
 965 </screen>
 966       <para>Lists all OSTs in the file system.</para>
 967       <screen>
 968 $ lfs osts
 969 </screen>
 970       <para>Lists space usage per OST and MDT in human-readable format.</para>
 971       <screen>
 972 $ lfs df -h
 973 </screen>
 974       <para>Lists inode usage per OST and MDT.</para>
 975       <screen>
 976 $ lfs df -i
 977 </screen>
 978       <para>List space or inode usage for a specific OST pool.</para>
 979       <screen>
 980 $ lfs df --pool
 981 <replaceable>filesystem</replaceable>[.
 982 <replaceable>pool</replaceable>] |
 983 <replaceable>pathname</replaceable>
 984 </screen>
 985       <para>List quotas of user 'bob'.</para>
 986       <screen>
 987 $ lfs quota -u bob /mnt/lustre
 988 </screen>
 989       <para>Show grace times for user quotas on
 990       <literal>/mnt/lustre</literal>.</para>
 991       <screen>
 992 $ lfs quota -t -u /mnt/lustre
 993 </screen>
 994       <para>Changes file owner and group.</para>
 995       <screen>
 996 $ lfs quotachown -i /mnt/lustre
 997 </screen>
 998       <para>Checks quotas for user and group. Turns on quotas after making the
 999       check.</para>
1000       <screen>
1001 $ lfs quotacheck -ug /mnt/lustre
1002 </screen>
1003       <para>Turns on quotas of user and group.</para>
1004       <screen>
1005 $ lfs quotaon -ug /mnt/lustre
1006 </screen>
1007       <para>Turns off quotas of user and group.</para>
1008       <screen>
1009 $ lfs quotaoff -ug /mnt/lustre
1010 </screen>
1011       <para>Sets quotas of user 'bob', with a 1 GB block quota hardlimit and a
1012       2 GB block quota softlimit.</para>
1013       <screen>
1014 $ lfs setquota -u bob --block-softlimit 2000000 --block-hardlimit 1000000
1015 /mnt/lustre
1016 </screen>
1017       <para>Sets grace times for user quotas: 1000 seconds for block quotas, 1
1018       week and 4 days for inode quotas.</para>
1019       <screen>
1020 $ lfs setquota -t -u --block-grace 1000 --inode-grace 1w4d /mnt/lustre
1021 </screen>
1022       <para>Checks the status of all servers (MDT, OST)</para>
1023       <screen>
1024 $ lfs check servers
1025 </screen>
1026       <para>Creates a file striped on two OSTs from the pool
1027       <literal>my_pool</literal></para>
1028       <screen>
1029 $ lfs setstripe --pool my_pool -c 2 /mnt/lustre/file
1030 </screen>
1031       <para>Lists the pools defined for the mounted Lustre file system
1032       <literal>/mnt/lustre</literal></para>
1033       <screen>
1034 $ lfs pool_list /mnt/lustre/
1035 </screen>
1036       <para>Lists the OSTs which are members of the pool
1037       <literal>my_pool</literal> in file system
1038       <literal>my_fs</literal></para>
1039       <screen>
1040 $ lfs pool_list my_fs.my_pool
1041 </screen>
1042       <para>Finds all directories/files associated with
1043       <literal>poolA</literal>.</para>
1044       <screen>
1045 $ lfs find /mnt/lustre --pool poolA
1046 </screen>
1047       <para>Finds all directories/files not associated with a pool.</para>
1048       <screen>
1049 $ lfs find /mnt//lustre --pool ""
1050 </screen>
1051       <para>Finds all directories/files associated with pool.</para>
1052       <screen>
1053 $ lfs find /mnt/lustre ! --pool ""
1054 </screen>
1055       <para>Associates a directory with the pool
1056       <literal>my_pool</literal>, so all new files and directories are created
1057       in the pool.</para>
1058       <screen>
1059 $ lfs setstripe --pool my_pool /mnt/lustre/dir
1060 </screen>
1061     </section>
1062     <section remap="h5">
1063       <title>See Also</title>
1064       <para>
1065         <xref linkend="dbdoclet.50438219_38274" />
1066       </para>
1067     </section>
1068   </section>
1069   <section xml:id="dbdoclet.50438206_42260">
1070     <title>
1071       <indexterm>
1072         <primary>lfs_migrate</primary>
1073       </indexterm>
1074       <literal>lfs_migrate</literal>
1075     </title>
1076     <para>The
1077     <literal>lfs_migrate</literal> utility is a simple tool to migrate files
1078     between Lustre OSTs.</para>
1079     <section remap="h5">
1080       <title>Synopsis</title>
1081       <screen>
1082 lfs_migrate [-c <replaceable>stripecount</replaceable>] [-h] [-l] [-n] [-q] [-R] [-s] [-y]
1083 [file|directory ...]
1084 </screen>
1085     </section>
1086     <section remap="h5">
1087       <title>Description</title>
1088       <para>The
1089       <literal>lfs_migrate</literal> utility is a simple tool to assist
1090       migration of files between Lustre OSTs. The utility copies each specified
1091       file to a new file, verifies the file contents have not changed, and then
1092       renames the new file to the original filename. This allows balanced space
1093       usage between OSTs, moving files off OSTs that are starting to show
1094       hardware problems (though are still functional) or OSTs that will be
1095       discontinued.</para>
1096       <warning>
1097         <para>For versions of Lustre before 2.5,
1098         <literal>lfs_migrate</literal> is not closely integrated with the MDS,
1099         it cannot determine whether a file is currently open and/or in-use by
1100         other applications or nodes. This makes it UNSAFE for use on files that
1101         might be modified by other applications, since the migrated file is
1102         only a copy of the current file. This results in the old file becoming
1103         an open-unlinked file and any modifications to that file are
1104         lost.</para>
1105       </warning>
1106       <para>Files to be migrated can be specified as command-line arguments. If
1107       a directory is specified on the command-line then all files within the
1108       directory are migrated. If no files are specified on the command-line,
1109       then a list of files is read from the standard input, making
1110       <literal>lfs_migrate</literal> suitable for use with
1111       <literal>lfs</literal> find to locate files on specific OSTs and/or
1112       matching other file attributes.</para>
1113       <para>The current file allocation policies on the MDS dictate where the
1114       new files are placed, taking into account whether specific OSTs have been
1115       disabled on the MDS via
1116       <literal>lctl</literal>(preventing new files from being allocated there),
1117       whether some OSTs are overly full (reducing the number of files placed on
1118       those OSTs), or if there is a specific default file striping for the
1119       target directory (potentially changing the stripe count, stripe size, OST
1120       pool, or OST index of a new file).</para>
1121       <note>
1122         <para>The
1123         <literal>lfs_migrate</literal> utility can also be used in some cases to
1124         reduce file
1125         <indexterm>
1126           <primary>fragmentation</primary>
1127         </indexterm>fragmentation. File fragmentation will typically reduce
1128         Lustre file system performance. File fragmentation may be observed on
1129         an aged file system and will commonly occur if the file was written by
1130         many threads. Provided there is sufficient free space (or if it was
1131         written when the file system was nearly full) that is less fragmented
1132         than the file being copied, re-writing a file with
1133         <literal>lfs_migrate</literal> will result in a migrated file with
1134         reduced fragmentation. The tool
1135         <literal>filefrag</literal> can be used to report file fragmentation.
1136         See
1137         <xref linkend="dbdoclet.50438206_75125" /></para>
1138       </note>
1139       <note>
1140         <para>As long as a file has extent lengths of tens of megabytes (
1141         <replaceable>read_bandwidth * seek_time</replaceable>) or more, the
1142         read performance for the file will not be significantly impacted by
1143         fragmentation, since the read pipeline can be filled by large reads
1144         from disk even with an occasional disk seek.</para>
1145       </note>
1146     </section>
1147     <section remap="h5">
1148       <title>Options</title>
1149       <para>Options supporting
1150       <literal>lfs_migrate</literal> are described below.</para>
1151       <informaltable frame="all">
1152         <tgroup cols="2">
1153           <colspec colname="c1" colwidth="50*" />
1154           <colspec colname="c2" colwidth="50*" />
1155           <thead>
1156             <row>
1157               <entry>
1158                 <para>
1159                   <emphasis role="bold">Option</emphasis>
1160                 </para>
1161               </entry>
1162               <entry>
1163                 <para>
1164                   <emphasis role="bold">Description</emphasis>
1165                 </para>
1166               </entry>
1167             </row>
1168           </thead>
1169           <tbody>
1170             <row>
1171               <entry>
1172                 <para>
1173                   <literal>-c
1174                   <replaceable>stripecount</replaceable></literal>
1175                 </para>
1176               </entry>
1177               <entry>
1178                 <para>Restripe file using the specified stripe count. This
1179                 option may not be specified at the same time as the
1180                 <literal>-R</literal> option.</para>
1181               </entry>
1182             </row>
1183             <row>
1184               <entry>
1185                 <para>
1186                   <literal>-h</literal>
1187                 </para>
1188               </entry>
1189               <entry>
1190                 <para>Display help information.</para>
1191               </entry>
1192             </row>
1193             <row>
1194               <entry>
1195                 <literal>-l</literal>
1196               </entry>
1197               <entry>
1198                 <para>Migrate files with hard links (skips, by default). Files
1199                 with multiple hard links are split into multiple separate files
1200                 by
1201                 <literal>lfs_migrate</literal>, so they are skipped, by
1202                 default, to avoid breaking the hard links.</para>
1203               </entry>
1204             </row>
1205             <row>
1206               <entry>
1207                 <literal>-n</literal>
1208               </entry>
1209               <entry>
1210                 <para>Only print the names of files to be migrated.</para>
1211               </entry>
1212             </row>
1213             <row>
1214               <entry>
1215                 <para>
1216                   <literal>-q</literal>
1217                 </para>
1218               </entry>
1219               <entry>
1220                 <para>Run quietly (does not print filenames or status).</para>
1221               </entry>
1222             </row>
1223             <row>
1224               <entry>
1225                 <literal>-R</literal>
1226               </entry>
1227               <entry>Restripe file using default directory striping instead of
1228               keeping striping. This option may not be specified at the same
1229               time as the
1230               <literal>-c</literal> option.</entry>
1231             </row>
1232             <row>
1233               <entry>
1234                 <literal>-s</literal>
1235               </entry>
1236               <entry>Skip file data comparison after migrate. Default is to
1237               compare migrated file against original to verify
1238               correctness.</entry>
1239             </row>
1240             <row>
1241               <entry>
1242                 <para>
1243                   <literal>-y</literal>
1244                 </para>
1245               </entry>
1246               <entry>
1247                 <para>Answer '
1248                 <literal>y</literal>' to usage warning without prompting (for
1249                 scripts, use with caution).</para>
1250               </entry>
1251             </row>
1252           </tbody>
1253         </tgroup>
1254       </informaltable>
1255     </section>
1256     <section remap="h5">
1257       <title>Examples</title>
1258       <para>Rebalances all files in
1259       <literal>/mnt/lustre/dir</literal>.</para>
1260       <screen>
1261 $ lfs_migrate /mnt/lustre/file
1262 </screen>
1263       <para>Migrates files in /test filesystem on OST004 larger than 4 GB in
1264       size.</para>
1265       <screen>
1266 $ lfs find /test -obd test-OST004 -size +4G | lfs_migrate -y
1267 </screen>
1268     </section>
1269     <section remap="h5">
1270       <title>See Also</title>
1271       <para>
1272         <xref linkend="dbdoclet.50438206_94597" />
1273       </para>
1274     </section>
1275   </section>
1276   <section xml:id="dbdoclet.50438206_75125">
1277     <title>
1278       <indexterm>
1279         <primary>filefrag</primary>
1280       </indexterm>
1281       <literal>filefrag</literal>
1282     </title>
1283     <para>The
1284     <literal>e2fsprogs</literal> package contains the
1285     <literal>filefrag</literal> tool which reports the extent of file
1286     fragmentation.</para>
1287     <section remap="h5">
1288       <title>Synopsis</title>
1289       <screen>
1290 filefrag [ -belsv ] [ files...  ]
1291 </screen>
1292     </section>
1293     <section remap="h5">
1294       <title>Description</title>
1295       <para>The
1296       <literal>filefrag</literal> utility reports the extent of fragmentation in
1297       a given file. The
1298       <literal>filefrag</literal> utility obtains the extent information from
1299       Lustre files using the
1300       <literal>FIEMAP ioctl</literal>, which is efficient and fast, even for
1301       very large files.</para>
1302       <para>In default mode
1303       <footnote>
1304         <para>The default mode is faster than the verbose/extent mode since it
1305         only counts the number of extents.</para>
1306       </footnote>,
1307       <literal>filefrag</literal> prints the number of physically discontiguous
1308       extents in the file. In extent or verbose mode, each extent is printed
1309       with details such as the blocks allocated on each OST. For a Lustre file
1310       system, the extents are printed in device offset order (i.e. all of the
1311       extents for one OST first, then the next OST, etc.), not file logical
1312       offset order. If the file logical offset order was used, the Lustre
1313       striping would make the output very verbose and difficult to see if there
1314       was file fragmentation or not.</para>
1315       <note>
1316         <para>Note that as long as a file has extent lengths of tens of
1317         megabytes or more (i.e.
1318         <replaceable>read_bandwidth * seek_time &gt;
1319         extent_length</replaceable>), the read performance for the file will
1320         not be significantly impacted by fragmentation, since file readahead
1321         can fully utilize the disk disk bandwidth even with occasional
1322         seeks.</para>
1323       </note>
1324       <para>In default mode
1325       <footnote>
1326         <para>The default mode is faster than the verbose/extent mode.</para>
1327       </footnote>,
1328       <literal>filefrag</literal> returns the number of physically discontiguous
1329       extents in the file. In extent or verbose mode, each extent is printed
1330       with details. For a Lustre file system, the extents are printed in device
1331       offset order, not logical offset order.</para>
1332     </section>
1333     <section remap="h5">
1334       <title>Options</title>
1335       <para>The options and descriptions for the
1336       <literal>filefrag</literal> utility are listed below.</para>
1337       <informaltable frame="all">
1338         <tgroup cols="2">
1339           <colspec colname="c1" colwidth="50*" />
1340           <colspec colname="c2" colwidth="50*" />
1341           <thead>
1342             <row>
1343               <entry>
1344                 <para>
1345                   <emphasis role="bold">Option</emphasis>
1346                 </para>
1347               </entry>
1348               <entry>
1349                 <para>
1350                   <emphasis role="bold">Description</emphasis>
1351                 </para>
1352               </entry>
1353             </row>
1354           </thead>
1355           <tbody>
1356             <row>
1357               <entry>
1358                 <para>
1359                   <literal>-b</literal>
1360                 </para>
1361               </entry>
1362               <entry>
1363                 <para>Uses the 1024-byte blocksize for the output. By default,
1364                 this blocksize is used by the Lustre file system, since OSTs
1365                 may use different block sizes.</para>
1366               </entry>
1367             </row>
1368             <row>
1369               <entry>
1370                 <para>
1371                   <literal>-e</literal>
1372                 </para>
1373               </entry>
1374               <entry>
1375                 <para>Uses the extent mode when printing the output. This is
1376                 the default for Lustre files in verbose mode.</para>
1377               </entry>
1378             </row>
1379             <row>
1380               <entry>
1381                 <para>
1382                   <literal>-l</literal>
1383                 </para>
1384               </entry>
1385               <entry>
1386                 <para>Displays extents in LUN offset order. This is the only
1387                 available mode for Lustre.</para>
1388               </entry>
1389             </row>
1390             <row>
1391               <entry>
1392                 <para>
1393                   <literal>-s</literal>
1394                 </para>
1395               </entry>
1396               <entry>
1397                 <para>Synchronizes any unwritten file data to disk before
1398                 requesting the mapping.</para>
1399               </entry>
1400             </row>
1401             <row>
1402               <entry>
1403                 <para>
1404                   <literal>-v</literal>
1405                 </para>
1406               </entry>
1407               <entry>
1408                 <para>Prints the file's layout in verbose mode when checking
1409                 file fragmentation, including the logical to physical mapping
1410                 for each extent in the file and the OST index.</para>
1411               </entry>
1412             </row>
1413           </tbody>
1414         </tgroup>
1415       </informaltable>
1416     </section>
1417     <section remap="h5">
1418       <title>Examples</title>
1419       <para>Lists default output.</para>
1420       <screen>
1421 $ filefrag /mnt/lustre/foo
1422 /mnt/lustre/foo: 13 extents found
1423 </screen>
1424       <para>Lists verbose output in extent format.</para>
1425       <screen>
1426 $ filefrag -v /mnt/lustre/foo
1427 Filesystem type is: bd00bd0
1428 File size of /mnt/lustre/foo is 1468297786 (1433888 blocks of 1024 bytes)
1429  ext:     device_logical:        physical_offset: length:  dev: flags:
1430    0:        0..  122879: 2804679680..2804802559: 122880: 0002: network
1431    1:   122880..  245759: 2804817920..2804940799: 122880: 0002: network
1432    2:   245760..  278527: 2804948992..2804981759:  32768: 0002: network
1433    3:   278528..  360447: 2804982784..2805064703:  81920: 0002: network
1434    4:   360448..  483327: 2805080064..2805202943: 122880: 0002: network
1435    5:   483328..  606207: 2805211136..2805334015: 122880: 0002: network
1436    6:   606208..  729087: 2805342208..2805465087: 122880: 0002: network
1437    7:   729088..  851967: 2805473280..2805596159: 122880: 0002: network
1438    8:   851968..  974847: 2805604352..2805727231: 122880: 0002: network
1439    9:   974848.. 1097727: 2805735424..2805858303: 122880: 0002: network
1440   10:  1097728.. 1220607: 2805866496..2805989375: 122880: 0002: network
1441   11:  1220608.. 1343487: 2805997568..2806120447: 122880: 0002: network
1442   12:  1343488.. 1433599: 2806128640..2806218751:  90112: 0002: network
1443 /mnt/lustre/foo: 13 extents found
1444 </screen>
1445     </section>
1446   </section>
1447   <section xml:id="dbdoclet.50438206_86244">
1448     <title>
1449       <indexterm>
1450         <primary>mount</primary>
1451       </indexterm>
1452       <literal>mount</literal>
1453     </title>
1454     <para>The standard
1455     <literal>mount(8)</literal> Linux command is used to mount a Lustre file
1456     system. When mounting a Lustre file system, mount(8) executes the
1457     <literal>/sbin/mount.lustre</literal> command to complete the mount. The
1458     mount command supports these options specific to a Lustre file
1459     system:</para>
1460     <informaltable frame="all">
1461       <tgroup cols="2">
1462         <colspec colname="c1" colwidth="50*" />
1463         <colspec colname="c2" colwidth="50*" />
1464         <thead>
1465           <row>
1466             <entry>
1467               <para>
1468                 <emphasis role="bold">Server options</emphasis>
1469               </para>
1470             </entry>
1471             <entry>
1472               <para>
1473                 <emphasis role="bold">Description</emphasis>
1474               </para>
1475             </entry>
1476           </row>
1477         </thead>
1478         <tbody>
1479           <row>
1480             <entry>
1481               <para>
1482                 <literal>abort_recov</literal>
1483               </para>
1484             </entry>
1485             <entry>
1486               <para>Aborts recovery when starting a target</para>
1487             </entry>
1488           </row>
1489           <row>
1490             <entry>
1491               <para>
1492                 <literal>nosvc</literal>
1493               </para>
1494             </entry>
1495             <entry>
1496               <para>Starts only MGS/MGC servers</para>
1497             </entry>
1498           </row>
1499           <row>
1500             <entry>
1501               <para>
1502                 <literal>nomgs</literal>
1503               </para>
1504             </entry>
1505             <entry>
1506               <para>Start a MDT with a co-located MGS without starting the
1507               MGS</para>
1508             </entry>
1509           </row>
1510           <row>
1511             <entry>
1512               <para>
1513                 <literal>exclude</literal>
1514               </para>
1515             </entry>
1516             <entry>
1517               <para>Starts with a dead OST</para>
1518             </entry>
1519           </row>
1520           <row>
1521             <entry>
1522               <para>
1523                 <literal>md_stripe_cache_size</literal>
1524               </para>
1525             </entry>
1526             <entry>
1527               <para>Sets the stripe cache size for server side disk with a
1528               striped raid configuration</para>
1529             </entry>
1530           </row>
1531         </tbody>
1532       </tgroup>
1533     </informaltable>
1534     <informaltable frame="all">
1535       <tgroup cols="2">
1536         <colspec colname="c1" colwidth="50*" />
1537         <colspec colname="c2" colwidth="50*" />
1538         <thead>
1539           <row>
1540             <entry>
1541               <para>
1542                 <emphasis role="bold">Client options</emphasis>
1543               </para>
1544             </entry>
1545             <entry>
1546               <para>
1547                 <emphasis role="bold">Description</emphasis>
1548               </para>
1549             </entry>
1550           </row>
1551         </thead>
1552         <tbody>
1553           <row>
1554             <entry>
1555               <para>
1556                 <literal>flock/noflock/localflock</literal>
1557               </para>
1558             </entry>
1559             <entry>
1560               <para>Enables/disables global flock or local flock support</para>
1561             </entry>
1562           </row>
1563           <row>
1564             <entry>
1565               <para>
1566                 <literal>user_xattr/nouser_xattr</literal>
1567               </para>
1568             </entry>
1569             <entry>
1570               <para>Enables/disables user-extended attributes</para>
1571             </entry>
1572           </row>
1573           <row>
1574             <entry>
1575               <para>
1576                 <literal>user_fid2path/nouser_fid2path</literal>
1577               </para>
1578             </entry>
1579             <entry>
1580               <para condition="l23">Enables/disables FID to path translation by
1581               regular users</para>
1582             </entry>
1583           </row>
1584           <row>
1585             <entry>
1586               <para>
1587                 <literal>retry=</literal>
1588               </para>
1589             </entry>
1590             <entry>
1591               <para>Number of times a client will retry to mount the file
1592               system</para>
1593             </entry>
1594           </row>
1595         </tbody>
1596       </tgroup>
1597     </informaltable>
1598   </section>
1599   <section xml:id="dbdoclet.50438206_56217">
1600     <title>Handling Timeouts</title>
1601     <para>Timeouts are the most common cause of hung applications. After a
1602     timeout involving an MDS or failover OST, applications attempting to access
1603     the disconnected resource wait until the connection gets
1604     established.</para>
1605     <para>When a client performs any remote operation, it gives the server a
1606     reasonable amount of time to respond. If a server does not reply either due
1607     to a down network, hung server, or any other reason, a timeout occurs which
1608     requires a recovery.</para>
1609     <para>If a timeout occurs, a message (similar to this one), appears on the
1610     console of the client, and in
1611     <literal>/var/log/messages</literal>:</para>
1612     <screen>
1613 LustreError: 26597:(client.c:810:ptlrpc_expire_one_request()) @@@ timeout
1614
1615 req@a2d45200 x5886/t0 o38-&gt;mds_svc_UUID@NID_mds_UUID:12 lens 168/64 ref 1 fl
1616
1617 RPC:/0/0 rc 0
1618 </screen>
1619   </section>
1620 </chapter>