Whamcloud - gitweb
LUDOC-521 pcc: fix mistake in example
[doc/manual.git] / ManagingLNet.xml
index ea3fc20..40189ff 100644 (file)
@@ -1,25 +1,28 @@
-<?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="managinglnet">
+<?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="managinglnet">
   <title xml:id="managinglnet.title">Managing Lustre Networking (LNet)</title>
   <para>This chapter describes some tools for managing Lustre networking (LNet) and includes the
     following sections:</para>
   <itemizedlist>
     <listitem>
-      <para><xref linkend="dbdoclet.50438203_51732"/></para>
+      <para><xref linkend="updating_health_status_of_router"/></para>
     </listitem>
     <listitem>
-      <para><xref linkend="dbdoclet.50438203_48703"/></para>
+      <para><xref linkend="starting_stoping_lnet"/></para>
     </listitem>
     <listitem>
-      <para><xref linkend="dbdoclet.50438203_82542"/></para>
+      <para><xref linkend="multi_rail_configuration_lnet"/></para>
     </listitem>
     <listitem>
-      <para><xref linkend="dbdoclet.50438203_78227"/></para>
+      <para><xref linkend="load_balancing_ib_network"/></para>
     </listitem>
     <listitem>
       <para><xref linkend="managinglnet.configuringroutes"/></para>
     </listitem>
   </itemizedlist>
-  <section xml:id="dbdoclet.50438203_51732">
+  <section xml:id="updating_health_status_of_router">
       <title><indexterm><primary>LNet</primary><secondary>management</secondary></indexterm>
           Updating the Health Status of a Peer or Router</title>
     <para>There are two mechanisms to update the health status of a peer or a router:</para>
@@ -44,7 +47,7 @@
       </listitem>
     </itemizedlist>
   </section>
-  <section xml:id="dbdoclet.50438203_48703">
+  <section xml:id="starting_stoping_lnet">
       <title><indexterm><primary>LNet</primary><secondary>starting/stopping</secondary></indexterm>Starting and Stopping LNet</title>
     <para>The Lustre software automatically starts and stops LNet, but it can also be manually
       started in a standalone manner. This is particularly useful to verify that your networking
@@ -92,36 +95,47 @@ To remove all Lustre modules, run:</para>
       </note>
     </section>
   </section>
-  <section xml:id="dbdoclet.50438203_82542">
-    <title><indexterm><primary>LNet</primary><secondary>multi-rail configuration</secondary></indexterm>Multi-Rail Configurations with LNet</title>
-    <para>To aggregate bandwidth across both rails of a dual-rail IB cluster (o2iblnd) <footnote>
-        <para>Multi-rail configurations are only supported by o2iblnd; other IB LNDs do not support multiple interfaces.</para>
+  <section xml:id="multi_rail_configuration_lnet">
+    <title><indexterm><primary>LNet</primary><secondary>hardware multi-rail
+    configuration</secondary></indexterm>Hardware Based Multi-Rail
+    Configurations with LNet</title>
+    <para>To aggregate bandwidth across both rails of a dual-rail IB cluster
+        (o2iblnd) <footnote>
+        <para>Hardware multi-rail configurations are only supported by o2iblnd;
+        other IB LNDs do not support multiple interfaces.</para>
       </footnote> using LNet, consider these points:</para>
     <itemizedlist>
       <listitem>
-        <para>LNet can work with multiple rails, however, it does not load balance across them. The actual rail used for any communication is determined by the peer NID.</para>
+        <para>LNet can work with multiple rails, however, it does not load
+        balance across them. The actual rail used for any communication is
+        determined by the peer NID.</para>
       </listitem>
       <listitem>
-        <para>Multi-rail LNet configurations do not provide an additional level of network fault
-          tolerance. The configurations described below are for bandwidth aggregation only. </para>
+        <para>Hardware multi-rail LNet configurations do not provide an
+        additional level of network fault tolerance. The configurations
+        described below are for bandwidth aggregation only.</para>
       </listitem>
       <listitem>
-        <para>A Lustre node always uses the same local NID to communicate with a given peer NID. The criteria used to determine the local NID are:</para>
+        <para>A Lustre node always uses the same local NID to communicate with a
+        given peer NID. The criteria used to determine the local NID are:</para>
         <itemizedlist>
-          <listitem>
-            <para condition='l25'>Lowest route priority number (lower number, higher priority).</para>
-          </listitem>
+             <listitem>
+               <para condition='l25'>Lowest route priority number (lower number,
+            higher priority).</para>
+             </listitem>
           <listitem>
             <para>Fewest hops (to minimize routing), and</para>
           </listitem>
           <listitem>
-            <para>Appears first in the &quot;<literal>networks</literal>&quot; or &quot;<literal>ip2nets</literal>&quot; LNet configuration strings</para>
+            <para>Appears first in the &quot;<literal>networks</literal>&quot;
+            or &quot;<literal>ip2nets</literal>&quot; LNet configuration strings
+            </para>
           </listitem>
         </itemizedlist>
       </listitem>
     </itemizedlist>
   </section>
-  <section xml:id="dbdoclet.50438203_78227">
+  <section xml:id="load_balancing_ib_network">
     <title><indexterm>
         <primary>LNet</primary>
         <secondary>InfiniBand load balancing</secondary>
@@ -233,7 +247,7 @@ ents&quot;</screen>
         &quot;even&quot; servers with <literal>o2ib2</literal> on <literal>rail1</literal>.</para>
     </section>
   </section>
-  <section xml:id="managinglnet.configuringroutes" condition='l24'>
+  <section xml:id="managinglnet.configuringroutes">
     <title><indexterm><primary>LNet</primary></indexterm>Dynamically Configuring
     LNet Routes</title>
     <para>Two scripts are provided:
@@ -263,7 +277,8 @@ ents&quot;</screen>
       script, since the script only adds extra routes after the router is
       identified.  To ensure that a router is identified correctly, make sure to
       add its local NID in the routes parameter in the modprobe lustre
-      configuration file.  See <xref linkend='dbdoclet.50438293_15350'/>.</para>
+      configuration file.
+      See <xref linkend='tuning_lnet_mod_params'/>.</para>
     </section>
     <section remap="h3">
       <title><indexterm><primary>LNet</primary></indexterm><literal>lustre_routes_conversion</literal></title>
@@ -288,3 +303,4 @@ tcp1: { gateway: 10.1.1.4@tcp0 }</literal></screen>
     </section>
   </section>
 </chapter>
+<!--vim:expandtab:shiftwidth=2:tabstop=8:-->