Author: jstephens
Date: 2007-01-25 09:20:38 +0100 (Thu, 25 Jan 2007)
New Revision: 4631
Log:
Documented Cluster bugfixes:
Cluster/General:
Bug #19956, Bug #19645, Bug #21858, Bug #22103, Bug #24820,
Bug #25364, Bug #25468, Bug #25554, Bug #25562, Bug #25567,
Bug #25587, Bug #25668, Bug #25686, Bug #25711, Bug #25746
Disk Data:
Bug #24667, Bug #25296, Bug #25636
Cluster Replication:
Bug #25329, Bug #25387, Bug #25755
Cluster APIs:
Bug #25487
Documented change in LockPagesInMainMemory parameter.
Modified:
branches/telcos/refman-5.1/mysql-cluster.xml
branches/telcos/refman-5.1/news-5.1.xml
trunk/refman-4.1/news-4.1.xml
trunk/refman-5.0/mysql-cluster.xml
trunk/refman-5.0/news-5.0.xml
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: branches/telcos/refman-5.1/mysql-cluster.xml
===================================================================
--- branches/telcos/refman-5.1/mysql-cluster.xml 2007-01-24 21:29:57 UTC (rev 4630)
+++ branches/telcos/refman-5.1/mysql-cluster.xml 2007-01-25 08:20:38 UTC (rev 4631)
Changed blocks: 2, Lines Added: 62, Lines Deleted: 20; 4351 bytes
@@ -4894,26 +4894,68 @@
<itemizedlist>
<listitem>
- <indexterm>
-
<primary><literal>LockPagesInMainMemory</literal></primary>
- </indexterm>
+
+
+ <indexterm>
+
<primary><literal>LockPagesInMainMemory</literal></primary>
+ </indexterm>
+
+ <para id="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">
+ <literal>LockPagesInMainMemory</literal>
+ </para>
+
+
+ <para>
+ For a number of operating systems, including Solaris and
+ Linux, it is possible to lock a process into memory and so
+ avoid any swapping to disk. This can be used to help
+ guarantee the cluster's real-time characteristics.
+ </para>
+
+ <para>
+ Beginning with MySQL 5.1.14/NDB-6.1.1, this parameter takes one of
+ the integer values <literal>0</literal>,
+ <literal>1</literal>, or <literal>2</literal>,
which act
+ as follows:
+ </para>
+
+ <itemizedlist>
+ <listitem>
+ <para>
+ <literal>0</literal>: Disables locking. This is the
+ default value.
+ </para>
+ </listitem>
+ <listitem>
+ <para>
+ <literal>1</literal>: Performs the lock after
+ allocating memory for the process.
+ </para>
+ </listitem>
+ <listitem>
+ <para>
+ <literal>2</literal>: Performs the lock before
+ memory for the process is allocated.
+ </para>
+ </listitem>
+ </itemizedlist>
+
+ <para>
+ Previously, this parameter was a Boolean.
+ <literal>0</literal> or <literal>false</literal>
was the
+ default setting, and disabled lcocking.
+ <literal>1</literal> or <literal>true</literal>
enabled
+ locking of the process after its memory was allocated.
+ <emphasis role="bold">Important</emphasis>: Beginning with
+ MySQL 5.1.14/NDB-6.1.1, it is no longer possible to use
+ <literal>true</literal> or
<literal>false</literal> for
+ the value of this parameter; when upgrading from a
+ previous version, you must change the value to
+ <literal>0</literal>, <literal>1</literal>, or
+ <literal>2</literal>.
+ </para>
+ </listitem>
- <para id="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">
- <literal>LockPagesInMainMemory</literal>
- </para>
-
- <para>
- For a number of operating systems, including Solaris and
- Linux, it is possible to lock a process into memory and so
- avoid any swapping to disk. This can be used to help
- guarantee the cluster's real-time characteristics.
- </para>
-
- <para>
- This feature is disabled by default.
- </para>
- </listitem>
-
<listitem>
<indexterm>
<primary><literal>StopOnError</literal></primary>
@@ -7305,7 +7347,7 @@
</row>
<row>
<entry><literal><link
linkend="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">LockPagesInMainMemory</link></literal></entry>
- <entry>true|false
(<literal>1</literal>|<literal>0</literal>)</entry>
+ <entry><emphasis>As of MySQL
5.1.14/NDB-6.1.1</emphasis>: integer; <emphasis>previously</emphasis>:
true|false
(<literal>1</literal>|<literal>0</literal>)</entry>
<entry>0</entry>
<entry>0</entry>
<entry>1</entry>
Modified: branches/telcos/refman-5.1/news-5.1.xml
===================================================================
--- branches/telcos/refman-5.1/news-5.1.xml 2007-01-24 21:29:57 UTC (rev 4630)
+++ branches/telcos/refman-5.1/news-5.1.xml 2007-01-25 08:20:38 UTC (rev 4631)
Changed blocks: 3, Lines Added: 108, Lines Deleted: 2; 4513 bytes
@@ -13,7 +13,8 @@
]>
<section id="news-5-1-x">
- <title>Changes in release 5.1.x (Development)</title>
+ <title>Changes in release 5.1.<replaceable>X</replaceable>
+ (Telco Edition)</title>
<indexterm>
<primary>changes</primary>
@@ -27,7 +28,7 @@
<para>
For a full list of changes, please refer to the changelog sections
- for each individual 5.1.x release.
+ for each individual release.
</para>
<remark>
@@ -35,6 +36,111 @@
<filename>TEMPLATE.news.xml</filename> template in the
<filename>refman-common</filename> directory.
</remark>
+
+ <section id="news-5-1-14-ndb-6-1-1">
+
+ <title>Changes in release MySQL 5.1.14/NDB-6.1.1-beta (Not yet
+ released)</title>
+
+ <para role="release-level">
+ This is a new Beta development release, fixing recently discovered
+ bugs.
+ </para>
+
+ <para role="release-non-production">
+ <remark>
+ 3. REMOVE WHOLE PARAGRAPH IF THIS IS A PRODUCTION RELEASE.
+ </remark>
+
+ <emphasis role="bold">NOTE:</emphasis> This Beta release, as any
+ other pre-production release, should not be installed on
+ <emphasis>production</emphasis> level systems or systems with
+ critical data. It is good practice to back up your data before
+ installing any new version of software. Although MySQL has worked
+ very hard to ensure a high level of quality, protect your data by
+ making a backup as you would for any software beta release. Please
+ refer to our bug database at <ulink url="http://bugs.mysql.com/"/>
+ for more details about the individual bugs fixed in this version.
+ </para>
+
+ <para role="release-special-announcement">
+ <remark>
+ 5. REMOVE WHOLE PARAGRAPH IF THERE IS NOTHING SPECIAL TO
+ ANNOUNCE.
+ </remark>
+ </para>
+
+ <para>
+ This section documents all changes and bug fixes that have been
+ applied in the second MySQL Cluster Telco Edition Beta release
+ — also known as MySQL Cluster 5.1.14/NDB-6.1.1 — since
+ the release of MySQL Cluster 5.1.14/NDB-6.1.0.
+ </para>
+
+ <para>
+ Functionality added or changed:
+ </para>
+
+ <itemizedlist>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The
+ <literal>LockPagesInMainMemory</literal> configuration
+ parameter has changed its type and possible values. For more
+ information, see
+ <literal><link
linkend="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">LockPagesInMainMemory</link></literal>.
+ (Bug #25686)
+ </para>
+
+ <para>
+ <emphasis role="bold">Important</emphasis>: The values
+ <literal>true</literal> and <literal>false</literal>
are no
+ longer accepted for this parameter. If you were using this
+ parameter and had it set to <literal>false</literal> in a
+ previous release, you must change it to <literal>0</literal>.
+ If you had this parameter set to <literal>true</literal>, you
+ should instead use <literal>1</literal> to obtain the same
+ behavior as previously, or <literal>2</literal> to take
+ advantage of new functionality introduced with this release
+ described in the section cited above.
+ </para>
+ </listitem>
+
+ </itemizedlist>
+
+ <para>
+ Bugs fixed:
+ </para>
+
+ <itemizedlist>
+
+ <listitem>
+ <para>
+ <emphasis role="bold">Cluster Replication</emphasis>: Certain
+ errors in replication setups could lead to subsequent node
+ failures. (Bug #25755)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Hosts in clusters with a large number of nodes could
+ experience excessive CPU usage while obtaining configuration
+ data. (Bug #25711)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ When stopping and restarting multiple data nodes, the last
+ node to be restarted would sometimes hang in Phase 100. (Bug
+ #19645)
+ </para>
+ </listitem>
+ </itemizedlist>
+
+ </section>
<section id="news-5-1-14-ndb-6-1-0">
Modified: trunk/refman-4.1/news-4.1.xml
===================================================================
--- trunk/refman-4.1/news-4.1.xml 2007-01-24 21:29:57 UTC (rev 4630)
+++ trunk/refman-4.1/news-4.1.xml 2007-01-25 08:20:38 UTC (rev 4631)
Changed blocks: 1, Lines Added: 9, Lines Deleted: 1; 930 bytes
@@ -221,11 +221,19 @@
</para>
<itemizedlist>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: In some circumstances,
+ shutting down the cluster could cause connected
+ <command>mysqld</command> processes to crash. (Bug #25668)
+ </para>
+ </listitem>
<listitem>
<para>
A table created with the <literal>ROW_FORMAT = FIXED</literal>
- table option loses the option if an index is added or dropped
+ table option lost the option if an index was added or dropped
with <literal>CREATE INDEX</literal> or <literal>DROP
INDEX</literal>. (Bug #23404)
</para>
Modified: trunk/refman-5.0/mysql-cluster.xml
===================================================================
--- trunk/refman-5.0/mysql-cluster.xml 2007-01-24 21:29:57 UTC (rev 4630)
+++ trunk/refman-5.0/mysql-cluster.xml 2007-01-25 08:20:38 UTC (rev 4631)
Changed blocks: 2, Lines Added: 43, Lines Deleted: 3; 3185 bytes
@@ -4890,16 +4890,56 @@
<para id="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">
<literal>LockPagesInMainMemory</literal>
</para>
-
+
+
<para>
For a number of operating systems, including Solaris and
Linux, it is possible to lock a process into memory and so
avoid any swapping to disk. This can be used to help
guarantee the cluster's real-time characteristics.
</para>
+
+ <para>
+ Beginning with MySQL 5.0.36, this parameter takes one of
+ the integer values
+ <literal>0</literal>,<literal>1</literal>, or
+ <literal>2</literal>, which act as follows:
+ </para>
+
+ <itemizedlist>
+ <listitem>
+ <para>
+ <literal>0</literal>: Disables locking. This is the
+ default value.
+ </para>
+ </listitem>
+ <listitem>
+ <para>
+ <literal>1</literal>: Performs the lock after
+ allocating memory for the process.
+ </para>
+ </listitem>
+ <listitem>
+ <para>
+ <literal>2</literal>: Performs the lock before
+ memory for the process is allocated.
+ </para>
+ </listitem>
+ </itemizedlist>
<para>
- This feature is disabled by default.
+ Previously, this parameter was a Boolean.
+ <literal>0</literal> or <literal>false</literal>
was the
+ default setting, and disabled lcocking.
+ <literal>1</literal> or <literal>true</literal>
enabled
+ locking of the process after its memory was allocated.
+ <emphasis role="bold">Important</emphasis>: Beginning with
+ MySQL 5.0.36, it is no longer possible to use
+ <literal>true</literal> or <literal>false</literal>
for
+ the value of this parameter; when upgrading from a
+ previous version, you must change the value to
+ <literal>0</literal>, <literal>1</literal>, or
+ <literal>2</literal>.
</para>
</listitem>
@@ -7250,7 +7290,7 @@
</row>
<row>
<entry><literal><link
linkend="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">LockPagesInMainMemory</link></literal></entry>
- <entry>true|false
(<literal>1</literal>|<literal>0</literal>)</entry>
+ <entry><emphasis>As of MySQL 5.0.36</emphasis>:
integer; <emphasis>previously</emphasis>: true|false
(<literal>1</literal>|<literal>0</literal>)</entry>
<entry>0</entry>
<entry>0</entry>
<entry>1</entry>
Modified: trunk/refman-5.0/news-5.0.xml
===================================================================
--- trunk/refman-5.0/news-5.0.xml 2007-01-24 21:29:57 UTC (rev 4630)
+++ trunk/refman-5.0/news-5.0.xml 2007-01-25 08:20:38 UTC (rev 4631)
Changed blocks: 1, Lines Added: 0, Lines Deleted: 11; 938 bytes
@@ -3550,17 +3550,6 @@
<listitem>
<para>
- <literal>NDB Cluster</literal>:
<literal>SELECT</literal>
- statements with a <literal>BLOB</literal> or
- <literal>TEXT</literal> column in the selected column list and
- a <literal>WHERE</literal> condition including a primary key
- lookup on a <literal>VARCHAR</literal> primary key produced
- empty result sets. (Bug #19956)
- </para>
- </listitem>
-
- <listitem>
- <para>
<literal>NDB Cluster</literal> (NDBAPI): On big-endian
platforms, <literal>NdbOperation::write_attr()</literal> did
not update 32-bit fields correctly. (Bug #19537)
Modified: trunk/refman-5.0/releasenotes-es-5.0.xml
===================================================================
--- trunk/refman-5.0/releasenotes-es-5.0.xml 2007-01-24 21:29:57 UTC (rev 4630)
+++ trunk/refman-5.0/releasenotes-es-5.0.xml 2007-01-25 08:20:38 UTC (rev 4631)
Changed blocks: 4, Lines Added: 115, Lines Deleted: 0; 5077 bytes
@@ -74,6 +74,30 @@
#22646)
</para>
</listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The
+ <literal>LockPagesInMainMemory</literal> configuration
+ parameter has changed its type and possible values. For more
+ information, see
+ <literal><link
linkend="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">LockPagesInMainMemory</link></literal>.
+ (Bug #25686)
+ </para>
+
+ <para>
+ <emphasis role="bold">Important</emphasis>: The values
+ <literal>true</literal> and <literal>false</literal>
are no
+ longer accepted for this parameter. If you were using this
+ parameter and had it set to <literal>false</literal> in a
+ previous release, you must change it to <literal>0</literal>.
+ If you had this parameter set to <literal>true</literal>, you
+ should instead use <literal>1</literal> to obtain the same
+ behavior as previously, or <literal>2</literal> to take
+ advantage of new functionality introduced with this release
+ described in the section cited above.
+ </para>
+ </listitem>
</itemizedlist>
@@ -184,6 +208,81 @@
#24507)
</para>
</listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal> (Cluster APIs): Deletion of an
+ <literal>Ndb_cluster_connection</literal> object took a very
+ long time. (Bug #25487)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: It was not possible to create
+ an <literal>NDB</literal> table with a key on two
+ <literal>VARCHAR</literal> columns where both columns had a
+ storage length in excess of 256. (Bug #25746)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: In some circumstances,
+ shutting down the cluster could cause connected
+ <command>mysqld</command> processes to crash. (Bug #25668)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Memory allocations for
+ <literal>TEXT</literal> columns were calculated incorrectly,
+ resulting in space being wasted and other issues. (Bug #25562)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The failure of a master node
+ during a node restart could lead to a resource leak, causing
+ later node failures. (Bug #25554)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The
<command>ndb_mgm</command>
+ client did not report node shutdown events correctly in
+ certain cases. (Bug #22103)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: A node shutdown occurred if
+ the master failed during a commit. (Bug #25364)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Creating a non-unique index
+ with the <literal>USING HASH</literal> clause silently created
+ an ordered index instead of issuing a warning. (Bug #24820)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>:
<literal>SELECT</literal>
+ statements with a <literal>BLOB</literal> or
+ <literal>TEXT</literal> column in the selected column list and
+ a <literal>WHERE</literal> condition including a primary key
+ lookup on a <literal>VARCHAR</literal> primary key produced
+ empty result sets. (Bug #19956)
+ </para>
+ </listitem>
</itemizedlist>
@@ -431,6 +530,14 @@
#17044)
</para>
</listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Hosts in clusters with a large
+ number of nodes could experience excessive CPU usage while
+ obtaining configuration data. (Bug #25711)
+ </para>
+ </listitem>
<listitem>
<para>
@@ -1447,6 +1554,14 @@
<listitem>
<para>
+ <literal>NDB Cluster</literal>: A potential memory leak in the
+ <literal>NDB</literal> storage engine's handling of file
+ operations was uncovered. (Bug #21858)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
<literal>NDB Cluster</literal>: The
<literal>HELP</literal>
command in the Cluster management client now provides
command-specific help. For example, <literal>HELP
Modified: trunk/refman-5.1/mysql-cluster.xml
===================================================================
--- trunk/refman-5.1/mysql-cluster.xml 2007-01-24 21:29:57 UTC (rev 4630)
+++ trunk/refman-5.1/mysql-cluster.xml 2007-01-25 08:20:38 UTC (rev 4631)
Changed blocks: 2, Lines Added: 46, Lines Deleted: 6; 3430 bytes
@@ -4896,26 +4896,66 @@
</para>
<itemizedlist>
-
+
<listitem>
<indexterm>
<primary><literal>LockPagesInMainMemory</literal></primary>
</indexterm>
-
+
<para id="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">
<literal>LockPagesInMainMemory</literal>
</para>
-
+
+
<para>
For a number of operating systems, including Solaris and
Linux, it is possible to lock a process into memory and so
avoid any swapping to disk. This can be used to help
guarantee the cluster's real-time characteristics.
</para>
-
+
<para>
- This feature is disabled by default.
+ Beginning with MySQL 5.0.36, this parameter takes one of
+ the integer values <literal>0</literal>,
+ <literal>1</literal>, or <literal>2</literal>,
which act
+ as follows:
</para>
+
+ <itemizedlist>
+ <listitem>
+ <para>
+ <literal>0</literal>: Disables locking. This is the
+ default value.
+ </para>
+ </listitem>
+ <listitem>
+ <para>
+ <literal>1</literal>: Performs the lock after
+ allocating memory for the process.
+ </para>
+ </listitem>
+ <listitem>
+ <para>
+ <literal>2</literal>: Performs the lock before
+ memory for the process is allocated.
+ </para>
+ </listitem>
+ </itemizedlist>
+
+ <para>
+ Previously, this parameter was a Boolean.
+ <literal>0</literal> or <literal>false</literal>
was the
+ default setting, and disabled lcocking.
+ <literal>1</literal> or <literal>true</literal>
enabled
+ locking of the process after its memory was allocated.
+ <emphasis role="bold">Important</emphasis>: Beginning with
+ MySQL 5.0.36, it is no longer possible to use
+ <literal>true</literal> or <literal>false</literal>
for
+ the value of this parameter; when upgrading from a
+ previous version, you must change the value to
+ <literal>0</literal>, <literal>1</literal>, or
+ <literal>2</literal>.
+ </para>
</listitem>
<listitem>
@@ -7257,7 +7297,7 @@
</row>
<row>
<entry><literal><link
linkend="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">LockPagesInMainMemory</link></literal></entry>
- <entry>true|false
(<literal>1</literal>|<literal>0</literal>)</entry>
+ <entry><emphasis>As of MySQL 5.1.15</emphasis>:
integer; <emphasis>previously</emphasis>: true|false
(<literal>1</literal>|<literal>0</literal>)</entry>
<entry>0</entry>
<entry>0</entry>
<entry>1</entry>
Modified: trunk/refman-5.1/news-5.1.xml
===================================================================
--- trunk/refman-5.1/news-5.1.xml 2007-01-24 21:29:57 UTC (rev 4630)
+++ trunk/refman-5.1/news-5.1.xml 2007-01-25 08:20:38 UTC (rev 4631)
Changed blocks: 3, Lines Added: 191, Lines Deleted: 8; 7757 bytes
@@ -249,6 +249,31 @@
removed. (Bug #22687)
</para>
</listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The
+ <literal>LockPagesInMainMemory</literal> configuration
+ parameter has changed its type and possible values. For more
+ information, see
+ <literal><link
linkend="mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">LockPagesInMainMemory</link></literal>.
+ (Bug #25686)
+ </para>
+
+ <para>
+ <emphasis role="bold">Important</emphasis>: The values
+ <literal>true</literal> and <literal>false</literal>
are no
+ longer accepted for this parameter. If you were using this
+ parameter and had it set to <literal>false</literal> in a
+ previous release, you must change it to <literal>0</literal>.
+ If you had this parameter set to <literal>true</literal>, you
+ should instead use <literal>1</literal> to obtain the same
+ behavior as previously, or <literal>2</literal> to take
+ advantage of new functionality introduced with this release
+ described in the section cited above.
+ </para>
+ </listitem>
+
</itemizedlist>
@@ -787,6 +812,172 @@
with partitioned tables in general. (Bug #22634)
</para>
</listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal> (Cluster APIs): Deletion of an
+ <literal>Ndb_cluster_connection</literal> object took a very
+ long time. (Bug #25487)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal> (Replication): Certain errors
+ in replication setups could lead to subsequent node failures.
+ (Bug #25755)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal> (Replication): Connecting a
+ <command>mysqld</command> to a cluster where not all nodes
+ were running, starting the remaining cluster nodes, and then
+ disconnecting from the cluster caused the
+ <command>mysqld</command> process to crash. (Bug #25387)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal> (Replication): Connecting an
+ API node to the cluster during a node restart while performing
+ database operations could cause the restarting node to fail.
+ (Bug #25329)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal> (Disk Data): Following 3 or
+ more missed local checkpoints by a cluster node, a restart of
+ the node caused incorrect undo information to be used for Disk
+ Data tables. (Bug #25636)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal> (Disk Data): Issuing a
+ <literal>TRUNCATE</literal> statement on a Disk Data table
+ caused the table to become an in-memory table. (Bug #25296)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal> (Disk Data): Changing a column
+ specification on a Disk Data table caused the table to become
+ an in-memory table. (Bug #24667)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: It was not possible to create
+ an <literal>NDB</literal> table with a key on two
+ <literal>VARCHAR</literal> columns where both columns had a
+ storage length in excess of 256. (Bug #25746)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Hosts in clusters with a large
+ number of nodes could experience excessive CPU usage while
+ obtaining configuration data. (Bug #25711)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: In some circumstances,
+ shutting down the cluster could cause connected
+ <command>mysqld</command> processes to crash. (Bug #25668)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Non-32-bit, non-aligned
+ columns were not handled correctly in explicitly partitioned
+ <literal>NDB</literal> tables. (Bug #25587)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Some aggregate queries such as
+ <literal>SELECT COUNT(*)</literal> performed a table scan on
+ <literal>NDB</literal> tables rather than checking table
+ statistics, causing such queries to perform much more slowly
+ in MySQL Cluster 5.1 than in 5.0. (Bug #25567)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Memory allocations for
+ <literal>TEXT</literal> columns were calculated incorrectly,
+ resulting in space being wasted and other issues. (Bug #25562)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The failure of a master node
+ during a node restart could lead to a resource leak, causing
+ later node failures. (Bug #25554)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The failure of a node during a
+ local checkpoint could lead to other node failures. (Bug
+ #25468)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: A node shutdown occurred if
+ the master failed during a commit. (Bug #25364)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Creating a non-unique index
+ with the <literal>USING HASH</literal> clause silently created
+ an ordered index instead of issuing a warning. (Bug #24820)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The
<command>ndb_mgm</command>
+ client did not report node shutdown events correctly in
+ certain cases. (Bug #22103)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: A potential memory leak in the
+ <literal>NDB</literal> storage engine's handling of file
+ operations was uncovered. (Bug #21858)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: When stopping and restarting
+ multiple data nodes, the last node to be restarted would
+ sometimes hang in Phase 100. (Bug #19645)
+ </para>
+ </listitem>
<listitem>
<para>
@@ -7017,14 +7208,6 @@
<listitem>
<para>
- <literal>NDB Cluster</literal>: When stopping and restarting
- multiple data nodes, the last node to be restarted would
- sometimes hang in Phase 100. (Bug #19645)
- </para>
- </listitem>
-
- <listitem>
- <para>
<literal>NDB Cluster</literal>: The
<literal>DATA_LENGTH</literal> and
<literal>AVG_ROW_LENGTH</literal> columns of the
| Thread |
|---|
| • svn commit - mysqldoc@docsrva: r4631 - branches/telcos/refman-5.1 trunk/refman-4.1 trunk/refman-5.0 trunk/refman-5.1 | jon | 25 Jan |