ConfiguringLustre.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="configuringlustre">
   5   <title xml:id="configuringlustre.title">Configuring a Lustre File
   6   System</title>
   7   <para>This chapter shows how to configure a simple Lustre file system
   8   comprised of a combined MGS/MDT, an OST and a client. It includes:</para>
   9   <itemizedlist>
  10     <listitem>
  11       <para>
  12         <xref linkend="lustre_configure" />
  13       </para>
  14     </listitem>
  15     <listitem>
  16       <para>
  17         <xref linkend="lustre_configure_additional_options" />
  18       </para>
  19     </listitem>
  20   </itemizedlist>
  21   <section xml:id="lustre_configure">
  22     <title>
  23     <indexterm>
  24       <primary>Lustre</primary>
  25       <secondary>configuring</secondary>
  26     </indexterm>Configuring a Simple Lustre File System</title>
  27     <para>A Lustre file system can be set up in a variety of configurations by
  28     using the administrative utilities provided with the Lustre software. The
  29     procedure below shows how to configure a simple Lustre file system
  30     consisting of a combined MGS/MDS, one OSS with two OSTs, and a client. For
  31     an overview of the entire Lustre installation procedure, see
  32     <xref linkend="installoverview" />.</para>
  33     <para>This configuration procedure assumes you have completed the
  34     following:</para>
  35     <itemizedlist>
  36       <listitem>
  37         <para>
  38         <emphasis>
  39           <emphasis role="bold">Set up and configured your hardware</emphasis>
  40         </emphasis>. For more information about hardware requirements, see
  41         <xref linkend="settinguplustresystem" />.</para>
  42       </listitem>
  43       <listitem>
  44         <para>
  45         <emphasis role="bold">Downloaded and installed the Lustre
  46         software.</emphasis>For more information about preparing for and
  47         installing the Lustre software, see
  48         <xref linkend="installinglustre" />.</para>
  49       </listitem>
  50     </itemizedlist>
  51     <para>The following optional steps should also be completed, if needed,
  52     before the Lustre software is configured:</para>
  53     <itemizedlist>
  54       <listitem>
  55         <para>
  56         <emphasis>Set up a hardware or software RAID on block devices to be
  57         used as OSTs or MDTs.</emphasis>For information about setting up RAID,
  58         see the documentation for your RAID controller or
  59         <xref linkend="configuringstorage" />.</para>
  60       </listitem>
  61       <listitem>
  62         <para>
  63         <emphasis>Set up network interface bonding on Ethernet
  64         interfaces.</emphasis>For information about setting up network
  65         interface bonding, see
  66         <xref linkend="settingupbonding" />.</para>
  67       </listitem>
  68       <listitem>
  69         <para>
  70         <emphasis>Set</emphasis> lnet
  71         <emphasis>module parameters to specify how Lustre Networking (LNet) is
  72         to be configured to work with a Lustre file system and test the LNet
  73         configuration.</emphasis>LNet will, by default, use the first TCP/IP
  74         interface it discovers on a system. If this network configuration is
  75         sufficient, you do not need to configure LNet. LNet configuration is
  76         required if you are using InfiniBand or multiple Ethernet
  77         interfaces.</para>
  78       </listitem>
  79     </itemizedlist>
  80     <para>For information about configuring LNet, see
  81     <xref linkend="configuringlnet" />. For information about testing LNet, see
  82
  83     <xref linkend="lnetselftest" />.</para>
  84     <itemizedlist>
  85       <listitem>
  86         <para>
  87         <emphasis>Run the benchmark script
  88         <literal>sgpdd-survey</literal> to determine baseline performance of
  89         your hardware.</emphasis>Benchmarking your hardware will simplify
  90         debugging performance issues that are unrelated to the Lustre software
  91         and ensure you are getting the best possible performance with your
  92         installation. For information about running
  93         <literal>sgpdd-survey</literal>, see
  94         <xref linkend="benchmarkingtests" />.</para>
  95       </listitem>
  96     </itemizedlist>
  97     <note>
  98       <para>The
  99       <literal>sgpdd-survey</literal> script overwrites the device being tested
 100       so it must be run before the OSTs are configured.</para>
 101     </note>
 102     <para>To configure a simple Lustre file system, complete these
 103     steps:</para>
 104     <orderedlist>
 105       <listitem>
 106         <para>Create a combined MGS/MDT file system on a block device. On the
 107         MDS node, run:</para>
 108         <screen>
 109 mkfs.lustre --fsname=
 110 <replaceable>fsname</replaceable> --mgs --mdt --index=0
 111 <replaceable>/dev/block_device</replaceable>
 112 </screen>
 113         <para>The default file system name (
 114         <literal>fsname</literal>) is
 115         <literal>lustre</literal>.</para>
 116         <note>
 117           <para>If you plan to create multiple file systems, the MGS should be
 118           created separately on its own dedicated block device, by
 119           running:</para>
 120           <screen>
 121 mkfs.lustre --fsname=
 122 <replaceable>fsname</replaceable> --mgs
 123 <replaceable>/dev/block_device</replaceable>
 124 </screen>
 125           <para>See
 126           <xref linkend="lustre_configure_multiple_fs" />for more details.</para>
 127         </note>
 128       </listitem>
 129       <listitem xml:id="addmdtindex">
 130         <para>Optionally add in additional MDTs.</para>
 131         <screen>
 132 mkfs.lustre --fsname=
 133 <replaceable>fsname</replaceable> --mgsnode=
 134 <replaceable>nid</replaceable> --mdt --index=1
 135 <replaceable>/dev/block_device</replaceable>
 136 </screen>
 137         <note>
 138           <para>Up to 4095 additional MDTs can be added.</para>
 139         </note>
 140       </listitem>
 141       <listitem>
 142         <para>Mount the combined MGS/MDT file system on the block device. On
 143         the MDS node, run:</para>
 144         <screen>
 145 mount -t lustre
 146 <replaceable>/dev/block_device</replaceable>
 147 <replaceable>/mount_point</replaceable>
 148 </screen>
 149         <note>
 150           <para>If you have created an MGS and an MDT on separate block
 151           devices, mount them both.</para>
 152         </note>
 153       </listitem>
 154       <listitem xml:id="format_ost">
 155         <para>Create the OST. On the OSS node, run:</para>
 156         <screen>
 157 mkfs.lustre --fsname=
 158 <replaceable>fsname</replaceable> --mgsnode=
 159 <replaceable>MGS_NID</replaceable> --ost --index=
 160 <replaceable>OST_index</replaceable>
 161 <replaceable>/dev/block_device</replaceable>
 162 </screen>
 163         <para>When you create an OST, you are formatting a
 164         <literal>ldiskfs</literal> or
 165         <literal>ZFS</literal> file system on a block storage device like you
 166         would with any local file system.</para>
 167         <para>You can have as many OSTs per OSS as the hardware or drivers
 168         allow. For more information about storage and memory requirements for a
 169         Lustre file system, see
 170         <xref linkend="settinguplustresystem" />.</para>
 171         <para>You can only configure one OST per block device. You should
 172         create an OST that uses the raw block device and does not use
 173         partitioning.</para>
 174         <para>You should specify the OST index number at format time in order
 175         to simplify translating the OST number in error messages or file
 176         striping to the OSS node and block device later on.</para>
 177         <para>If you are using block devices that are accessible from multiple
 178         OSS nodes, ensure that you mount the OSTs from only one OSS node at at
 179         time. It is strongly recommended that multiple-mount protection be
 180         enabled for such devices to prevent serious data corruption. For more
 181         information about multiple-mount protection, see
 182         <xref linkend="managingfailover" />.</para>
 183         <note>
 184           <para>The Lustre software currently supports block devices up to 128
 185           TB on Red Hat Enterprise Linux 5 and 6 (up to 8 TB on other
 186           distributions). If the device size is only slightly larger that 16
 187           TB, it is recommended that you limit the file system size to 16 TB at
 188           format time. We recommend that you not place DOS partitions on top of
 189           RAID 5/6 block devices due to negative impacts on performance, but
 190           instead format the whole disk for the file system.</para>
 191         </note>
 192       </listitem>
 193       <listitem xml:id="mount_ost">
 194         <para>Mount the OST. On the OSS node where the OST was created,
 195         run:</para>
 196         <screen>
 197 mount -t lustre
 198 <replaceable>/dev/block_device</replaceable>
 199 <replaceable>/mount_point</replaceable>
 200 </screen>
 201         <note>
 202           <para>To create additional OSTs, repeat Step
 203           <xref linkend="format_ost" />and Step
 204           <xref linkend="mount_ost" />, specifying the
 205           next higher OST index number.</para>
 206         </note>
 207       </listitem>
 208       <listitem xml:id="mount_on_client">
 209         <para>Mount the Lustre file system on the client. On the client node,
 210         run:</para>
 211         <screen>
 212 mount -t lustre
 213 <replaceable>MGS_node</replaceable>:/
 214 <replaceable>fsname</replaceable>
 215 <replaceable>/mount_point</replaceable>
 216 </screen>
 217         <note>
 218           <para>To mount the filesystem on additional clients, repeat Step
 219           <xref linkend="mount_on_client" />.</para>
 220         </note>
 221         <note>
 222           <para>If you have a problem mounting the file system, check the
 223           syslogs on the client and all the servers for errors and also check
 224           the network settings. A common issue with newly-installed systems is
 225           that
 226           <literal>hosts.deny</literal> or firewall rules may prevent
 227           connections on port 988.</para>
 228         </note>
 229       </listitem>
 230       <listitem>
 231         <para>Verify that the file system started and is working correctly. Do
 232         this by running
 233         <literal>lfs df</literal>,
 234         <literal>dd</literal> and
 235         <literal>ls</literal> commands on the client node.</para>
 236       </listitem>
 237       <listitem>
 238         <para>
 239         <emphasis>(Optional)</emphasis>Run benchmarking tools to validate the
 240         performance of hardware and software layers in the cluster. Available
 241         tools include:</para>
 242         <itemizedlist>
 243           <listitem>
 244             <para>
 245             <literal>obdfilter-survey</literal>- Characterizes the storage
 246             performance of a Lustre file system. For details, see
 247             <xref linkend="benchmark.ost_perf" />.</para>
 248           </listitem>
 249           <listitem>
 250             <para>
 251             <literal>ost-survey</literal>- Performs I/O against OSTs to detect
 252             anomalies between otherwise identical disk subsystems. For details,
 253             see
 254             <xref linkend="benchmark.ost_io" />.</para>
 255           </listitem>
 256         </itemizedlist>
 257       </listitem>
 258     </orderedlist>
 259     <section remap="h3">
 260       <title>
 261       <indexterm>
 262         <primary>Lustre</primary>
 263         <secondary>configuring</secondary>
 264         <tertiary>simple example</tertiary>
 265       </indexterm>Simple Lustre Configuration Example</title>
 266       <para>To see the steps to complete for a simple Lustre file system
 267       configuration, follow this example in which a combined MGS/MDT and two
 268       OSTs are created to form a file system called
 269       <literal>temp</literal>. Three block devices are used, one for the
 270       combined MGS/MDS node and one for each OSS node. Common parameters used
 271       in the example are listed below, along with individual node
 272       parameters.</para>
 273       <informaltable frame="all">
 274         <tgroup cols="4">
 275           <colspec colname="c1" colwidth="2*" />
 276           <colspec colname="c2" colwidth="25*" />
 277           <colspec colname="c3" colwidth="25*" />
 278           <colspec colname="c4" colwidth="25*" />
 279           <thead>
 280             <row>
 281               <entry nameend="c2" namest="c1">
 282                 <para>
 283                   <emphasis role="bold">Common Parameters</emphasis>
 284                 </para>
 285               </entry>
 286               <entry>
 287                 <para>
 288                   <emphasis role="bold">Value</emphasis>
 289                 </para>
 290               </entry>
 291               <entry>
 292                 <para>
 293                   <emphasis role="bold">Description</emphasis>
 294                 </para>
 295               </entry>
 296             </row>
 297           </thead>
 298           <tbody>
 299             <row>
 300               <entry>
 301                 <para>&#160;</para>
 302               </entry>
 303               <entry>
 304                 <para>
 305                   <emphasis role="bold">MGS node</emphasis>
 306                 </para>
 307               </entry>
 308               <entry>
 309                 <para>
 310                   <literal>10.2.0.1@tcp0</literal>
 311                 </para>
 312               </entry>
 313               <entry>
 314                 <para>Node for the combined MGS/MDS</para>
 315               </entry>
 316             </row>
 317             <row>
 318               <entry>
 319                 <para>&#160;</para>
 320               </entry>
 321               <entry>
 322                 <para>
 323                   <emphasis role="bold">file system</emphasis>
 324                 </para>
 325               </entry>
 326               <entry>
 327                 <para>
 328                   <literal>temp</literal>
 329                 </para>
 330               </entry>
 331               <entry>
 332                 <para>Name of the Lustre file system</para>
 333               </entry>
 334             </row>
 335             <row>
 336               <entry>
 337                 <para>&#160;</para>
 338               </entry>
 339               <entry>
 340                 <para>
 341                   <emphasis role="bold">network type</emphasis>
 342                 </para>
 343               </entry>
 344               <entry>
 345                 <para>
 346                   <literal>TCP/IP</literal>
 347                 </para>
 348               </entry>
 349               <entry>
 350                 <para>Network type used for Lustre file system
 351                 <literal>temp</literal></para>
 352               </entry>
 353             </row>
 354           </tbody>
 355         </tgroup>
 356       </informaltable>
 357       <informaltable frame="all">
 358         <tgroup cols="4">
 359           <colspec colname="c1" colwidth="25*" />
 360           <colspec colname="c2" colwidth="25*" />
 361           <colspec colname="c3" colwidth="25*" />
 362           <colspec colname="c4" colwidth="25*" />
 363           <thead>
 364             <row>
 365               <entry nameend="c2" namest="c1">
 366                 <para>
 367                   <emphasis role="bold">Node Parameters</emphasis>
 368                 </para>
 369               </entry>
 370               <entry>
 371                 <para>
 372                   <emphasis role="bold">Value</emphasis>
 373                 </para>
 374               </entry>
 375               <entry>
 376                 <para>
 377                   <emphasis role="bold">Description</emphasis>
 378                 </para>
 379               </entry>
 380             </row>
 381           </thead>
 382           <tbody>
 383             <row>
 384               <entry nameend="c4" namest="c1">
 385                 <para>MGS/MDS node</para>
 386               </entry>
 387             </row>
 388             <row>
 389               <entry>
 390                 <para>&#160;</para>
 391               </entry>
 392               <entry>
 393                 <para>
 394                   <emphasis role="bold">MGS/MDS node</emphasis>
 395                 </para>
 396               </entry>
 397               <entry>
 398                 <para>
 399                   <literal>mdt0</literal>
 400                 </para>
 401               </entry>
 402               <entry>
 403                 <para>MDS in Lustre file system
 404                 <literal>temp</literal></para>
 405               </entry>
 406             </row>
 407             <row>
 408               <entry>
 409                 <para>&#160;</para>
 410               </entry>
 411               <entry>
 412                 <para>
 413                   <emphasis role="bold">block device</emphasis>
 414                 </para>
 415               </entry>
 416               <entry>
 417                 <para>
 418                   <literal>/dev/sdb</literal>
 419                 </para>
 420               </entry>
 421               <entry>
 422                 <para>Block device for the combined MGS/MDS node</para>
 423               </entry>
 424             </row>
 425             <row>
 426               <entry>
 427                 <para>&#160;</para>
 428               </entry>
 429               <entry>
 430                 <para>
 431                   <emphasis role="bold">mount point</emphasis>
 432                 </para>
 433               </entry>
 434               <entry>
 435                 <para>
 436                   <literal>/mnt/mdt</literal>
 437                 </para>
 438               </entry>
 439               <entry>
 440                 <para>Mount point for the
 441                 <literal>mdt0</literal> block device (
 442                 <literal>/dev/sdb</literal>) on the MGS/MDS node</para>
 443               </entry>
 444             </row>
 445             <row>
 446               <entry nameend="c4" namest="c1">
 447                 <para>First OSS node</para>
 448               </entry>
 449             </row>
 450             <row>
 451               <entry>
 452                 <para>&#160;</para>
 453               </entry>
 454               <entry>
 455                 <para>
 456                   <emphasis role="bold">OSS node</emphasis>
 457                 </para>
 458               </entry>
 459               <entry>
 460                 <para>
 461                   <literal>oss0</literal>
 462                 </para>
 463               </entry>
 464               <entry>
 465                 <para>First OSS node in Lustre file system
 466                 <literal>temp</literal></para>
 467               </entry>
 468             </row>
 469             <row>
 470               <entry>
 471                 <para>&#160;</para>
 472               </entry>
 473               <entry>
 474                 <para>
 475                   <emphasis role="bold">OST</emphasis>
 476                 </para>
 477               </entry>
 478               <entry>
 479                 <para>
 480                   <literal>ost0</literal>
 481                 </para>
 482               </entry>
 483               <entry>
 484                 <para>First OST in Lustre file system
 485                 <literal>temp</literal></para>
 486               </entry>
 487             </row>
 488             <row>
 489               <entry>
 490                 <para>&#160;</para>
 491               </entry>
 492               <entry>
 493                 <para>
 494                   <emphasis role="bold">block device</emphasis>
 495                 </para>
 496               </entry>
 497               <entry>
 498                 <para>
 499                   <literal>/dev/sdc</literal>
 500                 </para>
 501               </entry>
 502               <entry>
 503                 <para>Block device for the first OSS node (
 504                 <literal>oss0</literal>)</para>
 505               </entry>
 506             </row>
 507             <row>
 508               <entry>
 509                 <para>&#160;</para>
 510               </entry>
 511               <entry>
 512                 <para>
 513                   <emphasis role="bold">mount point</emphasis>
 514                 </para>
 515               </entry>
 516               <entry>
 517                 <para>
 518                   <literal>/mnt/ost0</literal>
 519                 </para>
 520               </entry>
 521               <entry>
 522                 <para>Mount point for the
 523                 <literal>ost0</literal> block device (
 524                 <literal>/dev/sdc</literal>) on the
 525                 <literal>oss1</literal> node</para>
 526               </entry>
 527             </row>
 528             <row>
 529               <entry nameend="c4" namest="c1">
 530                 <para>Second OSS node</para>
 531               </entry>
 532             </row>
 533             <row>
 534               <entry>
 535                 <para></para>
 536               </entry>
 537               <entry>
 538                 <para>
 539                   <emphasis role="bold">OSS node</emphasis>
 540                 </para>
 541               </entry>
 542               <entry>
 543                 <para>
 544                   <literal>oss1</literal>
 545                 </para>
 546               </entry>
 547               <entry>
 548                 <para>Second OSS node in Lustre file system
 549                 <literal>temp</literal></para>
 550               </entry>
 551             </row>
 552             <row>
 553               <entry>
 554                 <para></para>
 555               </entry>
 556               <entry>
 557                 <para>
 558                   <emphasis role="bold">OST</emphasis>
 559                 </para>
 560               </entry>
 561               <entry>
 562                 <para>
 563                   <literal>ost1</literal>
 564                 </para>
 565               </entry>
 566               <entry>
 567                 <para>Second OST in Lustre file system
 568                 <literal>temp</literal></para>
 569               </entry>
 570             </row>
 571             <row>
 572               <entry />
 573               <entry>
 574                 <para>
 575                   <emphasis role="bold">block device</emphasis>
 576                 </para>
 577               </entry>
 578               <entry>
 579                 <para>
 580                   <literal>/dev/sdd</literal>
 581                 </para>
 582               </entry>
 583               <entry>
 584                 <para>Block device for the second OSS node (oss1)</para>
 585               </entry>
 586             </row>
 587             <row>
 588               <entry>
 589                 <para></para>
 590               </entry>
 591               <entry>
 592                 <para>
 593                   <emphasis role="bold">mount point</emphasis>
 594                 </para>
 595               </entry>
 596               <entry>
 597                 <para>
 598                   <literal>/mnt/ost1</literal>
 599                 </para>
 600               </entry>
 601               <entry>
 602                 <para>Mount point for the
 603                 <literal>ost1</literal> block device (
 604                 <literal>/dev/sdd</literal>) on the
 605                 <literal>oss1</literal> node</para>
 606               </entry>
 607             </row>
 608             <row>
 609               <entry nameend="c4" namest="c1">
 610                 <para>Client node</para>
 611               </entry>
 612             </row>
 613             <row>
 614               <entry>
 615                 <para></para>
 616               </entry>
 617               <entry>
 618                 <para>
 619                   <emphasis role="bold">client node</emphasis>
 620                 </para>
 621               </entry>
 622               <entry>
 623                 <para>
 624                   <literal>client1</literal>
 625                 </para>
 626               </entry>
 627               <entry>
 628                 <para>Client in Lustre file system
 629                 <literal>temp</literal></para>
 630               </entry>
 631             </row>
 632             <row>
 633               <entry>
 634                 <para></para>
 635               </entry>
 636               <entry>
 637                 <para>
 638                   <emphasis role="bold">mount point</emphasis>
 639                 </para>
 640               </entry>
 641               <entry>
 642                 <para>
 643                   <literal>/lustre</literal>
 644                 </para>
 645               </entry>
 646               <entry>
 647                 <para>Mount point for Lustre file system
 648                 <literal>temp</literal> on the
 649                 <literal>client1</literal> node</para>
 650               </entry>
 651             </row>
 652           </tbody>
 653         </tgroup>
 654       </informaltable>
 655       <note>
 656         <para>We recommend that you use 'dotted-quad' notation for IP addresses
 657         rather than host names to make it easier to read debug logs and debug
 658         configurations with multiple interfaces.</para>
 659       </note>
 660       <para>For this example, complete the steps below:</para>
 661       <orderedlist>
 662         <listitem>
 663           <para>Create a combined MGS/MDT file system on the block device. On
 664           the MDS node, run:</para>
 665           <screen>
 666 [root@mds /]# mkfs.lustre --fsname=temp --mgs --mdt --index=0 /dev/sdb
 667 </screen>
 668           <para>This command generates this output:</para>
 669           <screen>
 670     Permanent disk data:
 671 Target:            temp-MDT0000
 672 Index:             0
 673 Lustre FS: temp
 674 Mount type:        ldiskfs
 675 Flags:             0x75
 676    (MDT MGS first_time update )
 677 Persistent mount opts: errors=remount-ro,iopen_nopriv,user_xattr
 678 Parameters: mdt.identity_upcall=/usr/sbin/l_getidentity
 679
 680 checking for existing Lustre data: not found
 681 device size = 16MB
 682 2 6 18
 683 formatting backing filesystem ldiskfs on /dev/sdb
 684    target name             temp-MDTffff
 685    4k blocks               0
 686    options                 -i 4096 -I 512 -q -O dir_index,uninit_groups -F
 687 mkfs_cmd = mkfs.ext2 -j -b 4096 -L temp-MDTffff  -i 4096 -I 512 -q -O
 688 dir_index,uninit_groups -F /dev/sdb
 689 Writing CONFIGS/mountdata
 690 </screen>
 691         </listitem>
 692         <listitem>
 693           <para>Mount the combined MGS/MDT file system on the block device. On
 694           the MDS node, run:</para>
 695           <screen>
 696 [root@mds /]# mount -t lustre /dev/sdb /mnt/mdt
 697 </screen>
 698           <para>This command generates this output:</para>
 699           <screen>
 700 Lustre: temp-MDT0000: new disk, initializing
 701 Lustre: 3009:0:(lproc_mds.c:262:lprocfs_wr_identity_upcall()) temp-MDT0000:
 702 group upcall set to /usr/sbin/l_getidentity
 703 Lustre: temp-MDT0000.mdt: set parameter identity_upcall=/usr/sbin/l_getidentity
 704 Lustre: Server temp-MDT0000 on device /dev/sdb has started
 705 </screen>
 706         </listitem>
 707         <listitem xml:id="create_and_mount_ost">
 708           <para>Create and mount
 709           <literal>ost0</literal>.</para>
 710           <para>In this example, the OSTs (
 711           <literal>ost0</literal> and
 712           <literal>ost1</literal>) are being created on different OSS nodes (
 713           <literal>oss0</literal> and
 714           <literal>oss1</literal> respectively).</para>
 715           <orderedlist>
 716             <listitem>
 717               <para>Create
 718               <literal>ost0</literal>. On
 719               <literal>oss0</literal> node, run:</para>
 720               <screen>
 721 [root@oss0 /]# mkfs.lustre --fsname=temp --mgsnode=10.2.0.1@tcp0 --ost
 722 --index=0 /dev/sdc
 723 </screen>
 724               <para>The command generates this output:</para>
 725               <screen>
 726     Permanent disk data:
 727 Target:            temp-OST0000
 728 Index:             0
 729 Lustre FS: temp
 730 Mount type:        ldiskfs
 731 Flags:             0x72
 732 (OST first_time update)
 733 Persistent mount opts: errors=remount-ro,extents,mballoc
 734 Parameters: mgsnode=10.2.0.1@tcp
 735
 736 checking for existing Lustre data: not found
 737 device size = 16MB
 738 2 6 18
 739 formatting backing filesystem ldiskfs on /dev/sdc
 740    target name             temp-OST0000
 741    4k blocks               0
 742    options                 -I 256 -q -O dir_index,uninit_groups -F
 743 mkfs_cmd = mkfs.ext2 -j -b 4096 -L temp-OST0000  -I 256 -q -O
 744 dir_index,uninit_groups -F /dev/sdc
 745 Writing CONFIGS/mountdata
 746 </screen>
 747             </listitem>
 748             <listitem>
 749               <para>Mount ost0 on the OSS on which it was created. On
 750               <literal>oss0</literal> node, run:</para>
 751               <screen>
 752 root@oss0 /] mount -t lustre /dev/sdc /mnt/ost0
 753 </screen>
 754               <para>The command generates this output:</para>
 755               <screen>
 756 LDISKFS-fs: file extents enabled
 757 LDISKFS-fs: mballoc enabled
 758 Lustre: temp-OST0000: new disk, initializing
 759 Lustre: Server temp-OST0000 on device /dev/sdb has started
 760 </screen>
 761               <para>Shortly afterwards, this output appears:</para>
 762               <screen>
 763 Lustre: temp-OST0000: received MDS connection from 10.2.0.1@tcp0
 764 Lustre: MDS temp-MDT0000: temp-OST0000_UUID now active, resetting orphans
 765 </screen>
 766             </listitem>
 767           </orderedlist>
 768         </listitem>
 769         <listitem>
 770           <para>Create and mount
 771           <literal>ost1</literal>.</para>
 772           <orderedlist>
 773             <listitem>
 774               <para>Create ost1. On
 775               <literal>oss1</literal> node, run:</para>
 776               <screen>
 777 [root@oss1 /]# mkfs.lustre --fsname=temp --mgsnode=10.2.0.1@tcp0 \
 778            --ost --index=1 /dev/sdd
 779 </screen>
 780               <para>The command generates this output:</para>
 781               <screen>
 782     Permanent disk data:
 783 Target:            temp-OST0001
 784 Index:             1
 785 Lustre FS: temp
 786 Mount type:        ldiskfs
 787 Flags:             0x72
 788 (OST first_time update)
 789 Persistent mount opts: errors=remount-ro,extents,mballoc
 790 Parameters: mgsnode=10.2.0.1@tcp
 791
 792 checking for existing Lustre data: not found
 793 device size = 16MB
 794 2 6 18
 795 formatting backing filesystem ldiskfs on /dev/sdd
 796    target name             temp-OST0001
 797    4k blocks               0
 798    options                 -I 256 -q -O dir_index,uninit_groups -F
 799 mkfs_cmd = mkfs.ext2 -j -b 4096 -L temp-OST0001  -I 256 -q -O
 800 dir_index,uninit_groups -F /dev/sdc
 801 Writing CONFIGS/mountdata
 802 </screen>
 803             </listitem>
 804             <listitem>
 805               <para>Mount ost1 on the OSS on which it was created. On
 806               <literal>oss1</literal> node, run:</para>
 807               <screen>
 808 root@oss1 /] mount -t lustre /dev/sdd /mnt/ost1
 809 </screen>
 810               <para>The command generates this output:</para>
 811               <screen>
 812 LDISKFS-fs: file extents enabled
 813 LDISKFS-fs: mballoc enabled
 814 Lustre: temp-OST0001: new disk, initializing
 815 Lustre: Server temp-OST0001 on device /dev/sdb has started
 816 </screen>
 817               <para>Shortly afterwards, this output appears:</para>
 818               <screen>
 819 Lustre: temp-OST0001: received MDS connection from 10.2.0.1@tcp0
 820 Lustre: MDS temp-MDT0000: temp-OST0001_UUID now active, resetting orphans
 821 </screen>
 822             </listitem>
 823           </orderedlist>
 824         </listitem>
 825         <listitem>
 826           <para>Mount the Lustre file system on the client. On the client node,
 827           run:</para>
 828           <screen>
 829 root@client1 /] mount -t lustre 10.2.0.1@tcp0:/temp /lustre
 830 </screen>
 831           <para>This command generates this output:</para>
 832           <screen>
 833 Lustre: Client temp-client has started
 834 </screen>
 835         </listitem>
 836         <listitem>
 837           <para>Verify that the file system started and is working by running
 838           the
 839           <literal>df</literal>,
 840           <literal>dd</literal> and
 841           <literal>ls</literal> commands on the client node.</para>
 842           <orderedlist>
 843             <listitem>
 844               <para>Run the
 845               <literal>lfs df -h</literal> command:</para>
 846               <screen>
 847 [root@client1 /] lfs df -h
 848 </screen>
 849               <para>The
 850               <literal>lfs df -h</literal> command lists space usage per OST and
 851               the MDT in human-readable format. This command generates output
 852               similar to this:</para>
 853               <screen>
 854 UUID               bytes      Used      Available   Use%    Mounted on
 855 temp-MDT0000_UUID  8.0G      400.0M       7.6G        0%      /lustre[MDT:0]
 856 temp-OST0000_UUID  800.0G    400.0M     799.6G        0%      /lustre[OST:0]
 857 temp-OST0001_UUID  800.0G    400.0M     799.6G        0%      /lustre[OST:1]
 858 filesystem summary:  1.6T    800.0M       1.6T        0%      /lustre
 859 </screen>
 860             </listitem>
 861             <listitem>
 862               <para>Run the
 863               <literal>lfs df -ih</literal> command.</para>
 864               <screen>
 865 [root@client1 /] lfs df -ih
 866 </screen>
 867               <para>The
 868               <literal>lfs df -ih</literal> command lists inode usage per OST
 869               and the MDT. This command generates output similar to
 870               this:</para>
 871               <screen>
 872 UUID              Inodes      IUsed       IFree   IUse%     Mounted on
 873 temp-MDT0000_UUID   2.5M        32         2.5M      0%       /lustre[MDT:0]
 874 temp-OST0000_UUID   5.5M        54         5.5M      0%       /lustre[OST:0]
 875 temp-OST0001_UUID   5.5M        54         5.5M      0%       /lustre[OST:1]
 876 filesystem summary: 2.5M        32         2.5M      0%       /lustre
 877 </screen>
 878             </listitem>
 879             <listitem>
 880               <para>Run the
 881               <literal>dd</literal> command:</para>
 882               <screen>
 883 [root@client1 /] cd /lustre
 884 [root@client1 /lustre] dd if=/dev/zero of=/lustre/zero.dat bs=4M count=2
 885 </screen>
 886               <para>The
 887               <literal>dd</literal> command verifies write functionality by
 888               creating a file containing all zeros (
 889               <literal>0</literal>s). In this command, an 8 MB file is created.
 890               This command generates output similar to this:</para>
 891               <screen>
 892 2+0 records in
 893 2+0 records out
 894 8388608 bytes (8.4 MB) copied, 0.159628 seconds, 52.6 MB/s
 895 </screen>
 896             </listitem>
 897             <listitem>
 898               <para>Run the
 899               <literal>ls</literal> command:</para>
 900               <screen>
 901 [root@client1 /lustre] ls -lsah
 902 </screen>
 903               <para>The
 904               <literal>ls -lsah</literal> command lists files and directories in
 905               the current working directory. This command generates output
 906               similar to this:</para>
 907               <screen>
 908 total 8.0M
 909 4.0K drwxr-xr-x  2 root root 4.0K Oct 16 15:27 .
 910 8.0K drwxr-xr-x 25 root root 4.0K Oct 16 15:27 ..
 911 8.0M -rw-r--r--  1 root root 8.0M Oct 16 15:27 zero.dat
 912
 913 </screen>
 914             </listitem>
 915           </orderedlist>
 916         </listitem>
 917       </orderedlist>
 918       <para>Once the Lustre file system is configured, it is ready for
 919       use.</para>
 920     </section>
 921   </section>
 922   <section xml:id="lustre_configure_additional_options">
 923     <title>
 924     <indexterm>
 925       <primary>Lustre</primary>
 926       <secondary>configuring</secondary>
 927       <tertiary>additional options</tertiary>
 928     </indexterm>Additional Configuration Options</title>
 929     <para>This section describes how to scale the Lustre file system or make
 930     configuration changes using the Lustre configuration utilities.</para>
 931     <section remap="h3">
 932       <title>
 933       <indexterm>
 934         <primary>Lustre</primary>
 935         <secondary>configuring</secondary>
 936         <tertiary>for scale</tertiary>
 937       </indexterm>Scaling the Lustre File System</title>
 938       <para>A Lustre file system can be scaled by adding OSTs or clients. For
 939       instructions on creating additional OSTs repeat Step
 940       <xref linkend="create_and_mount_ost" />and Step
 941       <xref linkend="mount_ost" />above. For mounting
 942       additional clients, repeat Step
 943       <xref linkend="mount_on_client" />for each client.</para>
 944     </section>
 945     <section remap="h3">
 946       <title>
 947       <indexterm>
 948         <primary>Lustre</primary>
 949         <secondary>configuring</secondary>
 950         <tertiary>striping</tertiary>
 951       </indexterm>Changing Striping Defaults</title>
 952       <para>The default settings for the file layout stripe pattern are shown
 953       in
 954       <xref linkend="configuringlustre.tab.stripe" />.</para>
 955       <table frame="none" xml:id="configuringlustre.tab.stripe">
 956         <title>Default stripe pattern</title>
 957         <tgroup cols="3">
 958           <colspec colname="c1" colwidth="13*" />
 959           <colspec colname="c2" colwidth="13*" />
 960           <colspec colname="c3" colwidth="13*" />
 961           <tbody>
 962             <row>
 963               <entry>
 964                 <para>
 965                   <emphasis role="bold">File Layout Parameter</emphasis>
 966                 </para>
 967               </entry>
 968               <entry>
 969                 <para>
 970                   <emphasis role="bold">Default</emphasis>
 971                 </para>
 972               </entry>
 973               <entry>
 974                 <para>
 975                   <emphasis role="bold">Description</emphasis>
 976                 </para>
 977               </entry>
 978             </row>
 979             <row>
 980               <entry>
 981                 <para>
 982                   <literal>stripe_size</literal>
 983                 </para>
 984               </entry>
 985               <entry>
 986                 <para>1 MB</para>
 987               </entry>
 988               <entry>
 989                 <para>Amount of data to write to one OST before moving to the
 990                 next OST.</para>
 991               </entry>
 992             </row>
 993             <row>
 994               <entry>
 995                 <para>
 996                   <literal>stripe_count</literal>
 997                 </para>
 998               </entry>
 999               <entry>
1000                 <para>1</para>
1001               </entry>
1002               <entry>
1003                 <para>The number of OSTs to use for a single file.</para>
1004               </entry>
1005             </row>
1006             <row>
1007               <entry>
1008                 <para>
1009                   <literal>start_ost</literal>
1010                 </para>
1011               </entry>
1012               <entry>
1013                 <para>-1</para>
1014               </entry>
1015               <entry>
1016                 <para>The first OST where objects are created for each file.
1017                 The default -1 allows the MDS to choose the starting index
1018                 based on available space and load balancing.
1019                 <emphasis>It's strongly recommended not to change the default
1020                 for this parameter to a value other than -1.</emphasis></para>
1021               </entry>
1022             </row>
1023           </tbody>
1024         </tgroup>
1025       </table>
1026       <para>Use the
1027       <literal>lfs setstripe</literal> command described in
1028       <xref linkend="managingstripingfreespace" />to change the file layout
1029       configuration.</para>
1030     </section>
1031     <section remap="h3">
1032       <title>
1033       <indexterm>
1034         <primary>Lustre</primary>
1035         <secondary>configuring</secondary>
1036         <tertiary>utilities</tertiary>
1037       </indexterm>Using the Lustre Configuration Utilities</title>
1038       <para>If additional configuration is necessary, several configuration
1039       utilities are available:</para>
1040       <itemizedlist>
1041         <listitem>
1042           <para>
1043           <literal>mkfs.lustre</literal>- Use to format a disk for a Lustre
1044           service.</para>
1045         </listitem>
1046         <listitem>
1047           <para>
1048           <literal>tunefs.lustre</literal>- Use to modify configuration
1049           information on a Lustre target disk.</para>
1050         </listitem>
1051         <listitem>
1052           <para>
1053           <literal>lctl</literal>- Use to directly control Lustre features via
1054           an
1055           <literal>ioctl</literal> interface, allowing various configuration,
1056           maintenance and debugging features to be accessed.</para>
1057         </listitem>
1058         <listitem>
1059           <para>
1060           <literal>mount.lustre</literal>- Use to start a Lustre client or
1061           target service.</para>
1062         </listitem>
1063       </itemizedlist>
1064       <para>For examples using these utilities, see the topic
1065       <xref linkend="systemconfigurationutilities" /></para>
1066       <para>The
1067       <literal>lfs</literal> utility is useful for configuring and querying a
1068       variety of options related to files. For more information, see
1069       <xref linkend="userutilities" />.</para>
1070       <note>
1071         <para>Some sample scripts are included in the directory where the
1072         Lustre software is installed. If you have installed the Lustre source
1073         code, the scripts are located in the
1074         <literal>lustre/tests</literal> sub-directory. These scripts enable
1075         quick setup of some simple standard Lustre configurations.</para>
1076       </note>
1077     </section>
1078   </section>
1079 </chapter>
1080 <!--
1081   vim:expandtab:shiftwidth=2:tabstop=8:
1082   -->