1 <?xml version='1.0' encoding='UTF-8'?><chapter xmlns="http://docbook.org/ns/docbook" xmlns:xl="http://www.w3.org/1999/xlink" version="5.0" xml:lang="en-US" xml:id="upgradinglustre">
2 <title xml:id="upgradinglustre.title">Upgrading a Lustre File System</title>
3 <para>This chapter describes interoperability between Lustre software releases. It also provides
4 procedures for upgrading from a Lustre 1.8 release to a Lustre 2.x release, from a Lustre 2.x
5 release to a more recent Lustre 2.x release (major release upgrade), and from a Lustre 2.X.y
6 release to 2.X.y (minor release upgrade). It includes the following sections:</para>
9 <para><xref linkend="dbdoclet.50438205_82079"/></para>
12 <para><xref xmlns:xlink="http://www.w3.org/1999/xlink" linkend="Upgrading_2.x"/></para>
15 <para><xref xmlns:xlink="http://www.w3.org/1999/xlink" linkend="Upgrading_2.x.x"/></para>
18 <section xml:id="dbdoclet.50438205_82079">
20 <primary>Lustre</primary>
21 <secondary>upgrading</secondary>
23 </indexterm><indexterm>
24 <primary>upgrading</primary>
25 </indexterm>Release Interoperability and Upgrade Requirements</title>
26 <para><emphasis role="italic"><emphasis role="bold">Lustre software version 2.x release (major)
27 upgrade:</emphasis></emphasis><itemizedlist>
29 <para>All servers must be upgraded at the same time, while some or all clients may be
33 <para>All servers must be be upgraded to a Linux kernel supported by the Lustre software. See the Linux Test Matrix at <xref
34 xmlns:xlink="http://www.w3.org/1999/xlink" linkend="LustreTestMatrixTable"/> for a
35 list of tested Lustre distributions.</para>
38 <para>Clients to be upgraded to the Lustre 2.4 release or higher must be running a compatible Linux distribution. See the Linux Test Matrix at <xref
39 xmlns:xlink="http://www.w3.org/1999/xlink" linkend="LustreTestMatrixTable"/> for a
40 list of tested Linux distributions.</para>
42 </itemizedlist></para>
43 <para><emphasis role="italic"><emphasis role="bold">Lustre 2.X.y release (minor)
44 upgrade:</emphasis></emphasis></para>
47 <para>All servers must be upgraded at the same time, while some or all clients may be
51 <para>Rolling upgrades are supported for minor releases allowing individual servers and
52 clients to be upgraded without stopping the Lustre file system.</para>
56 <section xml:id="Upgrading_2.x">
58 <primary>upgrading</primary>
59 <secondary>major release (2.x to 2.x)</secondary>
60 </indexterm><indexterm>
61 <primary>wide striping</primary>
62 </indexterm><indexterm>
63 <primary>MDT</primary>
64 <secondary>multiple MDSx</secondary>
65 </indexterm><indexterm>
66 <primary>large_xattr</primary>
67 </indexterm><indexterm>
68 <primary>wide striping</primary>
69 <secondary>large_xattr</secondary>
70 </indexterm>Upgrading to Lustre Release 2.x (Major Release)</title>
71 <para> The procedure for upgrading a Lustre release 2.x file system to a more recent 2.x release
72 of the Lustre software is described in this section. </para>
74 <para>This procedure can also be used to upgrade Lustre release 1.8.6-wc1 or later to any
75 Lustre release 2.x. To upgrade other versions of 1.8.x, contact your support
79 <para condition="l22">In Lustre release 2.2, a feature has been added that allows striping
80 across up to 2000 OSTs. By default, this "wide striping" feature is disabled. It is
81 activated by setting the <literal>large_xattr</literal> option on the MDT using either
82 <literal>mkfs.lustre</literal> or <literal>tune2fs</literal>. For example after upgrading
83 an existing file system to Lustre release 2.2 or later, wide striping can be enabled by
84 running the following command on the MDT device before mounting
85 it:<screen>tune2fs -O large_xattr</screen>Once the wide striping feature is enabled and in
86 use on the MDT, it is not possible to directly downgrade the MDT file system to an earlier
87 version of the Lustre software that does not support wide striping. To disable wide striping:<orderedlist>
89 <para>Delete all wide-striped files. </para>
91 <para>Use <literal>lfs_migrate</literal> with the option <literal>-c</literal>
92 <replaceable>stripe_count</replaceable> (set <replaceable>stripe_count</replaceable>
93 to 160) to move the files to another location.</para>
96 <para>Unmount the MDT.</para>
99 <para>Run the following command to turn off the <literal>large_xattr</literal>
100 option:<screen>tune2fs -O ^large_xattr</screen></para>
104 <note condition="l23">
105 <para>To generate a list of all files with more than 160 stripes use <literal>lfs
106 find</literal> with the <literal>--stripe-count</literal>
107 option:<screen>lfs find ${mountpoint} --stripe-count=+160</screen></para>
109 <note condition="l24">
110 <para>In Lustre release 2.4, a new feature allows using multiple MDTs, which can each serve
111 one or more remote sub-directories in the file system. The <literal>root</literal> directory
112 is always located on MDT0. </para>
113 <para>Note that clients running a release prior to the Lustre 2.4 release can only see the
114 namespace hosted by MDT0 and will return an IO error if an attempt is made to access a
115 directory on another MDT.</para>
117 <para>To upgrade a Lustre release 2.x file system to a more recent major release, complete these
121 <para>Create a complete, restorable file system backup. </para>
123 <para>Before installing the Lustre software, back up ALL data. The Lustre software
124 contains kernel modifications that interact with storage devices and may introduce
125 security issues and data loss if not installed, configured, or administered properly. If
126 a full backup of the file system is not practical, a device-level backup of the MDT file
127 system is recommended. See <xref linkend="backupandrestore"/> for a procedure.</para>
131 <para>Shut down the file system by unmounting all clients and servers in the order shown
132 below (unmounting a block device causes the Lustre software to be shut down on that
134 <orderedlist numeration="loweralpha">
136 <para>Unmount the clients. On each client node, run:</para>
137 <screen>umount -a -t lustre</screen>
140 <para>Unmount the MDT. On the MDS node, run:</para>
141 <screen>umount -a -t lustre</screen>
144 <para>Unmount all the OSTs. On each OSS node, run:</para>
145 <screen>umount -a -t lustre</screen>
150 <para>Upgrade the Linux operating system on all servers to a compatible
151 (tested) Linux distribution and reboot. See the Linux Test Matrix at <xref
152 xmlns:xlink="http://www.w3.org/1999/xlink" linkend="LustreTestMatrixTable"/>.</para>
155 <para>Upgrade the Linux operating system on all clients to Red Hat Enterprise Linux 6 or
156 other compatible (tested) distribution and reboot. See the Linux Test Matrix at <xref
157 xmlns:xlink="http://www.w3.org/1999/xlink" linkend="LustreTestMatrixTable"/>.</para>
160 <para>Download the Lustre server RPMs for your platform from the <link
161 xl:href="https://wiki.hpdd.intel.com/display/PUB/Lustre+Releases">Lustre Releases</link>
162 repository. See <xref xmlns:xlink="http://www.w3.org/1999/xlink"
163 linkend="table_cnh_5m3_gk"/> for a list of required packages.</para>
166 <para>Install the Lustre server packages on all Lustre servers (MGS, MDSs, and OSSs).</para>
167 <orderedlist numeration="loweralpha">
169 <para>Log onto a Lustre server as the <literal>root</literal> user</para>
172 <para>Use the <literal>yum</literal> command to install the packages:</para>
174 <screen># yum --nogpgcheck install pkg1.rpm pkg2.rpm ...</screen>
178 <para>Verify the packages are installed correctly:</para>
180 <screen>rpm -qa|egrep "lustre|wc"</screen>
184 <para>Repeat these steps on each Lustre server.</para>
189 <para>Download the Lustre client RPMs for your platform from the <link
190 xl:href="https://wiki.hpdd.intel.com/display/PUB/Lustre+Releases">Lustre Releases</link>
191 repository. See <xref xmlns:xlink="http://www.w3.org/1999/xlink"
192 linkend="table_j3r_ym3_gk"/> for a list of required packages.</para>
194 <para>The version of the kernel running on a Lustre client must be the same as the version
195 of the <literal>lustre-client-modules-</literal><replaceable>ver</replaceable> package
196 being installed. If not, a compatible kernel must be installed on the client before the
197 Lustre client packages are installed.</para>
201 <para>Install the Lustre client packages on each of the Lustre clients to be
203 <orderedlist numeration="loweralpha">
205 <para>Log onto a Lustre client as the <literal>root</literal> user.</para>
208 <para>Use the <literal>yum</literal> command to install the packages:</para>
210 <screen># yum --nogpgcheck install pkg1.rpm pkg2.rpm ...</screen>
214 <para>Verify the packages were installed correctly:</para>
216 <screen># rpm -qa|egrep "lustre|kernel"</screen>
220 <para>Repeat these steps on each Lustre client.</para>
225 <para>(Optional) For upgrades to Lustre release 2.2 or higher, to enable wide striping on an
226 existing MDT, run the following command on the MDT
227 :<screen>mdt# tune2fs -O large_xattr <replaceable>device</replaceable></screen></para>
228 <para>For more information about wide striping, see <xref
229 xmlns:xlink="http://www.w3.org/1999/xlink" linkend="section_syy_gcl_qk"/>.</para>
232 <para>(Optional) For upgrades to Lustre release 2.4 or higher, to format an additional MDT,
233 complete these steps:<orderedlist numeration="loweralpha">
235 <para>Determine the index used for the first MDT (each MDT must have unique index).
236 Enter:<screen>client$ lctl dl | grep mdc
237 36 UP mdc lustre-MDT0000-mdc-ffff88004edf3c00
238 4c8be054-144f-9359-b063-8477566eb84e 5</screen></para>
239 <para>In this example, the next available index is 1.</para>
242 <para>Add the new block device as a new MDT at the next available index by entering
244 line):<screen>mds# mkfs.lustre --reformat --fsname=<replaceable>filesystem_name</replaceable> --mdt \
245 --mgsnode=<replaceable>mgsnode</replaceable> --index <replaceable>1</replaceable> <replaceable>/dev/mdt1_device</replaceable>
248 </orderedlist></para>
251 <para>(Optional) If you are upgrading to Lustre software release 2.3 or higher from Lustre
252 software version 2.2 or earlier and want to enable the quota feature, complete these
253 steps: <orderedlist numeration="loweralpha">
255 <para>Before setting up the file system, enter on both the MDS and
256 OSTs:<screen>tunefs.lustre --quota</screen></para>
259 <para>When setting up the file system,
260 enter:<screen>conf_param $FSNAME.quota.mdt=$QUOTA_TYPE
261 conf_param $FSNAME.quota.ost=$QUOTA_TYPE</screen></para>
263 </orderedlist></para>
266 <para>(Optional) If you are upgrading from Lustre release 1.8, you must manually enable the
267 FID-in-dirent feature. On the MDS,
268 enter:<screen>tune2fs –O dirdata /dev/<replaceable>mdtdev</replaceable></screen></para>
270 <para>This step is not reversible. Do not complete this step until you are sure you will
271 not be downgrading the Lustre software.</para>
273 <para>This step only enables FID-in-dirent for newly created files. If you are upgrading to
274 Lustre release 2.4, you can use LFSCK 1.5 to enable FID-in-dirent for existing files. For
275 more information about FID-in-dirent and related functionalities in LFSCK 1.5, see <xref
276 xmlns:xlink="http://www.w3.org/1999/xlink" linkend="understandinglustre.storageio"
280 <para>Start the Lustre file system by starting the components in the order shown in the
281 following steps:</para>
282 <orderedlist numeration="loweralpha">
284 <para>Mount the MGT. On the MGS, run<screen>mgs# mount -a -t lustre</screen></para>
287 <para>Mount the MDT(s). On each MDT, run:<screen>mds# mount -a -t lustre</screen></para>
290 <para>Mount all the OSTs. On each OSS node, run:</para>
291 <screen>oss# mount -a -t lustre</screen>
293 <para>This command assumes that all the OSTs are listed in the
294 <literal>/etc/fstab</literal> file. OSTs that are not listed in the
295 <literal>/etc/fstab</literal> file, must be mounted individually by running the
296 mount command:</para>
297 <screen> mount -t lustre <replaceable>/dev/block_device</replaceable> <replaceable>/mount_point</replaceable> </screen>
301 <para>Mount the file system on the clients. On each client node, run:</para>
302 <screen>client# mount -a -t lustre</screen>
308 <para>The mounting order described in the steps above must be followed for the intial mount
309 and registration of a Lustre file system after an upgrade. For a normal start of a Lustre
310 file system, the mounting order is MGT, OSTs, MDT(s), clients.</para>
312 <para>If you have a problem upgrading a Lustre file system, see <xref
313 xmlns:xlink="http://www.w3.org/1999/xlink" linkend="dbdoclet.50438198_30989"/> for some ways
316 <section xml:id="Upgrading_2.x.x">
318 <primary>upgrading</primary>
319 <secondary>2.X.y to 2.X.y (minor release)</secondary>
320 </indexterm>Upgrading to a Lustre Release 2.X.y (Minor Release)</title>
321 <para>Rolling upgrades are supported for upgrading from any Lustre release 2.X.y to a more
322 recent 2.X.y release. This allows the Lustre file system to continue to run while individual
323 servers (or their failover partners) and clients are upgraded one at a time. The procedure for
324 upgrading a Lustre release 2.X.y file system to a more recent minor release is described in
326 <para>To upgrade Lustre release 2.X.y to a more recent minor release, complete these
330 <para>Create a complete, restorable file system backup. </para>
332 <para>Before installing the Lustre software, back up ALL data. The Lustre software
333 contains kernel modifications that interact with storage devices and may introduce
334 security issues and data loss if not installed, configured, or administered properly. If
335 a full backup of the file system is not practical, a device-level backup of the MDT file
336 system is recommended. See <xref linkend="backupandrestore"/> for a procedure.</para>
340 <para>Download the Lustre server RPMs for your platform from the <link
341 xl:href="https://wiki.hpdd.intel.com/display/PUB/Lustre+Releases">Lustre Releases</link>
342 repository. See <xref xmlns:xlink="http://www.w3.org/1999/xlink"
343 linkend="table_cnh_5m3_gk"/> for a list of required packages.</para>
346 <para>For a rolling upgrade, complete any procedures required to keep the Lustre file system
347 running while the server to be upgraded is offline, such as failing over a primary server
348 to its secondary partner. </para>
351 <para>Unmount the Lustre server to be upgraded (MGS, MDS, or OSS)</para>
354 <para>Install the Lustre server packages on the Lustre server.</para>
355 <orderedlist numeration="loweralpha">
357 <para>Log onto the Lustre server as the <literal>root</literal> user</para>
360 <para>Use the <literal>yum</literal> command to install the packages:</para>
362 <screen># yum --nogpgcheck install pkg1.rpm pkg2.rpm ...</screen>
366 <para>Verify the packages are installed correctly:</para>
368 <screen>rpm -qa|egrep "lustre|wc"</screen>
372 <para>Mount the Lustre server to restart the Lustre software on the
373 server:<screen>server# mount -a -t lustre</screen></para>
376 <para>Repeat these steps on each Lustre server.</para>
381 <para>Download the Lustre client RPMs for your platform from the <link
382 xl:href="https://wiki.hpdd.intel.com/display/PUB/Lustre+Releases">Lustre Releases</link>
383 repository. See <xref xmlns:xlink="http://www.w3.org/1999/xlink"
384 linkend="table_j3r_ym3_gk"/> for a list of required packages.</para>
387 <para>Install the Lustre client packages on each of the Lustre clients to be
389 <orderedlist numeration="loweralpha">
391 <para>Log onto a Lustre client as the <literal>root</literal> user.</para>
394 <para>Use the <literal>yum</literal> command to install the packages:</para>
396 <screen># yum --nogpgcheck install pkg1.rpm pkg2.rpm ...</screen>
400 <para>Verify the packages were installed correctly:</para>
402 <screen># rpm -qa|egrep "lustre|kernel"</screen>
406 <para>Mount the Lustre client to restart the Lustre software on the
407 client:<screen>client# mount -a -t lustre</screen></para>
410 <para>Repeat these steps on each Lustre client.</para>
415 <para>If you have a problem upgrading a Lustre file system, see <xref
416 xmlns:xlink="http://www.w3.org/1999/xlink" linkend="dbdoclet.50438198_30989"/> for some
417 suggestions for how to get help.</para>