MySQL Lists are EOL. Please join:

List:Commits« Previous MessageNext Message »
From:jon Date:September 15 2007 1:21pm
Subject:svn commit - mysqldoc@docsrva: r7871 - in trunk: refman-5.0 refman-5.1
View as plain text  
Author: jstephens
Date: 2007-09-15 15:20:59 +0200 (Sat, 15 Sep 2007)
New Revision: 7871

Log:

Documented Cluster bugfixes:

  Bug #11313, Bug #20872, Bug #22240, Bug #23354, Bug #27494,
  Bug #27543, Bug #26793, Bug #27683, Bug #28253, Bug #28423,
  Bug #28683, Bug #28804, Bug #29228, Bug #29674, Bug #30024,
  Bug #30646, Bug #30667, Bug #30764, Bug #30914 (ClusterRep)

Also documented ndb_restore --no-upgrade|-u option



Modified:
   trunk/refman-5.0/releasenotes-es-5.0.xml
   trunk/refman-5.1/mysql-cluster.xml
   trunk/refman-5.1/news-5.1.xml


Modified: trunk/refman-5.0/releasenotes-es-5.0.xml
===================================================================
--- trunk/refman-5.0/releasenotes-es-5.0.xml	2007-09-14 22:16:13 UTC (rev 7870)
+++ trunk/refman-5.0/releasenotes-es-5.0.xml	2007-09-15 13:20:59 UTC (rev 7871)
Changed blocks: 3, Lines Added: 60, Lines Deleted: 2; 3135 bytes

@@ -56,7 +56,6 @@
       applied since the last MySQL Enterprise Server release (5.0.48).
     </para>
 
-<!--
     <para>
       Functionality added or changed:
     </para>

@@ -65,11 +64,21 @@
 
       <listitem>
         <para>
+          <literal>NDB Cluster</literal>: The output from the cluster
+          management client showing the progress of data node starts has
+          been improved. (Bug #23354)
         </para>
       </listitem>
 
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: Mapping of
+          <literal>NDB</literal> error codes to storage engine error
+          codes has been improved. (Bug #28423)
+        </para>
+      </listitem>
+
     </itemizedlist>
--->
 
     <para>
       Bugs fixed:

@@ -83,6 +92,55 @@
 
       <listitem>
         <para>
+          <literal>NDB Cluster</literal>: The description of the
+          <option>--print</option> option in the output from
+          <command>ndb_restore <option>--help</option></command> was
+          incorrect. (Bug #27683)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: An invalid subselect on an
+          <literal>NDB</literal> table could cause mysqld to crash. (Bug
+          #27494)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: Attempting to restore a backup
+          made on a cluster host using one endian to a machine using the
+          other endian could cause the cluster to fail. (Bug #29674)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: An attempt to perform a
+          <literal>SELECT ... FROM INFORMATION_SCHEMA.TABLES</literal>
+          whose result included information about <literal>NDB</literal>
+          tables for which the user had no privileges could crash the
+          MySQL Server on which the query was performed. (Bug #26793)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: A query using joins between
+          several large tables and requiring unique index lookups failed
+          to complete, eventually returning <errortext>Uknown
+          Error</errortext> after a very long period of time. This
+          occurred due to inadequate handling of instances where the
+          Transaction Coordinator ran out of
+          <literal>TransactionBufferMemory</literal>, when the cluster
+          should have returned NDB error code 4012 (<errortext>Request
+          ndbd time-out</errortext>). (Bug #28804)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
           Killing an SSL connection on platforms where MySQL is compiled
           with <literal>-DSIGNAL_WITH_VIO_CLOSE</literal> (Windows, Mac
           OS X, and some others) could crash the server. (Bug #28812)


Modified: trunk/refman-5.1/mysql-cluster.xml
===================================================================
--- trunk/refman-5.1/mysql-cluster.xml	2007-09-14 22:16:13 UTC (rev 7870)
+++ trunk/refman-5.1/mysql-cluster.xml	2007-09-15 13:20:59 UTC (rev 7871)
Changed blocks: 4, Lines Added: 26, Lines Deleted: 8; 2679 bytes

@@ -15362,6 +15362,20 @@
           </note>
 
           <para>
+            Formerly, when using <command>ndb_restore</command> to
+            restore a backup made from a MySQL 5.0 cluster to a 5.1
+            cluster, <literal>VARCHAR</literal> columns were not resized
+            and were recreated using the 5.0 fixed format. Beginning
+            with MySQL 5.1.19, <literal>ndb_restore</literal> recreates
+            such <literal>VARCHAR</literal> columns using MySQL Cluster
+            5.1&apos;s variable-width format. Also beginning with MySQL
+            5.1.19, this behavior can be overridden using the
+            <option>--no-upgrade</option> option (short form:
+            <option>-u</option>) when running
+            <command>ndb_restore</command>.
+          </para>
+
+          <para>
             Most of the options available for this program are shown in
             the following table:
           </para>

@@ -15479,6 +15493,13 @@
                     unless this option is used)</entry>
                 </row>
                 <row>
+                  <entry><option>--no-upgrade</option></entry>
+                  <entry><option>-u</option></entry>
+                  <entry>Do not re-create <literal>VARSIZE</literal> columns from a MySQL 5.0
+                    Cluster backup as variable-width columns (added in
+                    MySQL 5.1.19)</entry>
+                </row>
+                <row>
                   <entry><option>--nodeid</option></entry>
                   <entry><option>-n</option></entry>
                   <entry>Use backup files from node with the specified ID</entry>

@@ -18697,12 +18718,12 @@
 
             </itemizedlist>
           </para>
-          
-          &mccge-warning-begin;
+
           <para>
-            Beginning with MySQL 5.1.22-ndb-6.2.5, the behavior of this
-            utility and its available options have changed. (Bug #28683,
-            Bug #28253)
+            Beginning with MySQL 5.1.23
+            (<emphasis>&mccge-current;</emphasis>: 5.1.22-ndb-6.2.5),
+            the behavior of this utility and its available options have
+            changed. (Bug #28683, Bug #28253)
           </para>
 
           <para>

@@ -18820,9 +18841,6 @@
             specified using <option>--loadqueries</option>. Neither of
             these options has a default value.
           </para>
-          
-          
-          &mccge-warning-end-cluster;
 
         </refsection>
 


Modified: trunk/refman-5.1/news-5.1.xml
===================================================================
--- trunk/refman-5.1/news-5.1.xml	2007-09-14 22:16:13 UTC (rev 7870)
+++ trunk/refman-5.1/news-5.1.xml	2007-09-15 13:20:59 UTC (rev 7871)
Changed blocks: 5, Lines Added: 248, Lines Deleted: 2; 9362 bytes

@@ -67,6 +67,90 @@
         </para>
       </listitem>
 
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: The following improvements
+          have been made in the <command>ndb_size.pl</command> utility:
+
+          <itemizedlist>
+
+            <listitem>
+              <para>
+                The script is now able to be used with multiple
+                databases; lists of databases and tables can be excluded
+                from analysis.
+              </para>
+            </listitem>
+
+            <listitem>
+              <para>
+                Schema name information has been added to index table
+                calculations.
+              </para>
+            </listitem>
+
+            <listitem>
+              <para>
+                The database name is now an optional parameter, the
+                exclusion of which causes all databases to be examined.
+              </para>
+            </listitem>
+
+            <listitem>
+              <para>
+                If selecting from <literal>INFORMATION_SCHEMA</literal>
+                fails, the script now attempts to fall back to
+                <literal>SHOW TABLES</literal>.
+              </para>
+            </listitem>
+
+            <listitem>
+              <para>
+                A <option>--real_table_name</option> option has been
+                added; this designates a table to handle unique index
+                size calculations.
+              </para>
+            </listitem>
+
+            <listitem>
+              <para>
+                The report title has been amended to cover cases where
+                more than one database is being analyzed.
+              </para>
+            </listitem>
+
+          </itemizedlist>
+
+          (Bug #28683)
+        </para>
+
+        <para>
+          Support for a <option>--socket</option> option was also added.
+          (Bug #28253)
+        </para>
+
+        <para>
+          For more information, see
+          <xref linkend="mysql-cluster-utilities-ndb-size"/>.
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: The output from the cluster
+          management client showing the progress of data node starts has
+          been improved. (Bug #23354)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: Mapping of
+          <literal>NDB</literal> error codes to storage engine error
+          codes has been improved. (Bug #28423)
+        </para>
+      </listitem>
+
     </itemizedlist>
 
     <para>

@@ -81,6 +165,145 @@
 
       <listitem>
         <para>
+          <literal>NDB Cluster</literal>: The description of the
+          <option>--print</option> option in the output from
+          <command>ndb_restore <option>--help</option></command> was
+          incorrect. (Bug #27683)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: An insufficiently descriptive
+          and potentially misleading Error 4006 (<errortext>Connect
+          failure - out of connection objects...</errortext>) was
+          produced when either of the following two conditions occurred:
+
+          <orderedlist>
+
+            <listitem>
+              <para>
+                There were no more transaction records in the
+                transaction coordinator
+              </para>
+            </listitem>
+
+            <listitem>
+              <para>
+                an <literal>Ndb</literal> object in the NDB API was
+                initialized with insufficient parallellism
+              </para>
+            </listitem>
+
+          </orderedlist>
+
+          Separate error messages are now generated for each of these
+          two cases. (Bug #11313)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: An invalid subselect on an
+          <literal>NDB</literal> table could cause mysqld to crash. (Bug
+          #27494)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: Attempting to restore a backup
+          made on a cluster host using one endian to a machine using the
+          other endian could cause the cluster to fail. (Bug #29674)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: Restoring a backup made on a
+          cluster host using one endian to a machine using the other
+          endian failed for <literal>BLOB</literal> and
+          <literal>DATETIME</literal> columns. (Bug #27543, Bug #30024)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal> (Replication): Incorrect
+          handling of <literal>INSERT</literal> plus
+          <literal>DELETE</literal> operations with regard to local
+          checkpoints caused data node failures in multi-master
+          replication setups. (Bug #30914)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: When handling
+          <literal>BLOB</literal> columns, the addition of read locks to
+          the lock queue was not handled correctly. (Bug #30764)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: <command>ndb_size.pl</command>
+          failed on tables with <literal>FLOAT</literal> columns whose
+          definitions included commas (for example,
+          <literal>FLOAT(6,2)</literal>). (Bug #29228)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: Discovery of NDB tables did
+          not work correctly with <literal>INFORMATION_SCHEMA</literal>.
+          (Bug #30667)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: An attempt to perform a
+          <literal>SELECT ... FROM INFORMATION_SCHEMA.TABLES</literal>
+          whose result included information about <literal>NDB</literal>
+          tables for which the user had no privileges could crash the
+          MySQL Server on which the query was performed. (Bug #26793)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: An issue with the
+          <literal>mysql.ndb_apply_status</literal> table could cause
+          <literal>NDB</literal> schema autodiscovery to fail in certain
+          rare circumstances. (Bug #20872)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: A query using joins between
+          several large tables and requiring unique index lookups failed
+          to complete, eventually returning <errortext>Uknown
+          Error</errortext> after a very long period of time. This
+          occurred due to inadequate handling of instances where the
+          Transaction Coordinator ran out of
+          <literal>TransactionBufferMemory</literal>, when the cluster
+          should have returned NDB error code 4012 (<errortext>Request
+          ndbd time-out</errortext>). (Bug #28804)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: A filesystem close operation
+          could fail during a node or system restart. (Bug #30646)
+        </para>
+      </listitem>
+
+      <listitem>
+        <para>
           Killing an SSL connection on platforms where MySQL is compiled
           with <literal>-DSIGNAL_WITH_VIO_CLOSE</literal> (Windows, Mac
           OS X, and some others) could crash the server. (Bug #28812)

@@ -983,8 +1206,8 @@
 
         <listitem>
           <para>
-            A filesystem close could fail during a node or system
-            restart. (Bug #30646)
+            A filesystem close operation could fail during a node or
+            system restart. (Bug #30646)
           </para>
         </listitem>
 

@@ -996,6 +1219,15 @@
           </para>
         </listitem>
 
+        <listitem>
+          <para>
+            An issue with the <literal>mysql.ndb_apply_status</literal>
+            table could cause <literal>NDB</literal> schema
+            autodiscovery to fail in certain rare circumstances. (Bug
+            #20872)
+          </para>
+        </listitem>
+
       </itemizedlist>
 
     </section>

@@ -4656,6 +4888,20 @@
         </para>
       </listitem>
 
+      <listitem>
+        <para>
+          <literal>NDB Cluster</literal>: Formerly, restoring a cluster
+          backup made on a MySQL 5.0 Cluster to a 5.1 cluster using a
+          5.1 version of <command>ndb_restore</command> did not resize
+          <literal>VARCHAR</literal> columns as might be expected, but
+          now the default behavior of <command>ndb_restore</command> in
+          such cases is to resize the <literal>VARCHAR</literal>
+          columns. This new default behavior can be overridden using the
+          <option>--no-upgrade</option> (or <option>-u</option>) option
+          when invoking <command>ndb_restore</command>. (Bug #22240)
+        </para>
+      </listitem>
+
     </itemizedlist>
 
     <para>


Thread
svn commit - mysqldoc@docsrva: r7871 - in trunk: refman-5.0 refman-5.1jon15 Sep