Author: jstephens
Date: 2006-03-27 16:04:41 +0200 (Mon, 27 Mar 2006)
New Revision: 1685
Log:
Documenting Cluster bugfixes:
Bug #16772, Bug #17380, Bug #17431, Bug #17813, Bug #18118,
Bug #18295, Bug #18298, Bug #18352, Bug #18385, Bug #18414
Modified:
trunk/refman-5.0/ndbcluster.xml
trunk/refman-5.1/ndbcluster.xml
trunk/refman-common/news-4.1.xml
trunk/refman-common/news-5.0.xml
trunk/refman-common/news-5.1.xml
Modified: trunk/refman-5.0/ndbcluster.xml
===================================================================
--- trunk/refman-5.0/ndbcluster.xml 2006-03-27 13:48:57 UTC (rev 1684)
+++ trunk/refman-5.0/ndbcluster.xml 2006-03-27 14:04:41 UTC (rev 1685)
@@ -9127,12 +9127,18 @@
<listitem>
<para>
- <literal>INSERT IGNORE</literal> is supported only for
- primary keys, but not for unique keys. One possible
- workaround is to remove the constraint by dropping the
- unique index, perform any inserts, and then add the unique
- index again. (Bug #17431)
+ In MySQL 5.0.19 and earlier, <literal>INSERT IGNORE</literal>,
+ <literal>UPDATE IGNORE</literal>, and <literal>REPLACE</literal>
+ are supported only for primary keys, but not for unique keys. One
+ possible workaround is to remove the constraint by dropping the
+ unique index, perform any inserts, and then add the unique index
+ again.
</para>
+
+ <para>
+ This limitation is removed for <literal>INSERT IGNORE</literal>
+ and <literal>REPLACE</literal> in MySQL 5.0.20. (Bug #17431)
+ </para>
</listitem>
</itemizedlist>
Modified: trunk/refman-5.1/ndbcluster.xml
===================================================================
--- trunk/refman-5.1/ndbcluster.xml 2006-03-27 13:48:57 UTC (rev 1684)
+++ trunk/refman-5.1/ndbcluster.xml 2006-03-27 14:04:41 UTC (rev 1685)
@@ -11124,16 +11124,30 @@
However, spatial indexes are not supported.
</para>
</listitem>
-
+
<listitem>
<para>
- <literal>INSERT IGNORE</literal> is supported only for
- primary keys, but not for unique keys. One possible
- workaround is to remove the constraint by dropping the
- unique index, perform any inserts, and then add the unique
- index again. (Bug #17431)
+ <literal>CREATE TABLE</literal> statements may be no more than
+ 4096 characters in length. This limitation is lifted in MySQL
+ 5.1.9. (Bug #17813)
</para>
</listitem>
+
+ <listitem>
+ <para>
+ In MySQL 5.1.7 and earlier, <literal>INSERT IGNORE</literal>,
+ <literal>UPDATE IGNORE</literal>, and <literal>REPLACE</literal>
+ are supported only for primary keys, but not for unique keys. One
+ possible workaround is to remove the constraint by dropping the
+ unique index, perform any inserts, and then add the unique index
+ again.
+ </para>
+
+ <para>
+ This limitation is removed for <literal>INSERT IGNORE</literal>
+ and <literal>REPLACE</literal> in MySQL 5.1.8. (Bug #17431)
+ </para>
+ </listitem>
<listitem>
<para>
Modified: trunk/refman-common/news-4.1.xml
===================================================================
--- trunk/refman-common/news-4.1.xml 2006-03-27 13:48:57 UTC (rev 1684)
+++ trunk/refman-common/news-4.1.xml 2006-03-27 14:04:41 UTC (rev 1685)
@@ -190,12 +190,64 @@
</para>
<itemizedlist>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: A timeout in the handling of
+ an <literal>ABORT</literal> condition with more that 32
+ operations could yield a node failure. (Bug #18414)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: A node restart immediately
+ following a <literal>CREATE TABLE</literal> would fail.
+ <emphasis role="bold">Important</emphasis>: This fix supports
+ 2-node Clusters only. (Bug #18385)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: In event of a node failure
+ during a rollback, a <quote>false</quote> lock could be
+ established on the backup for that node, which lock could not
+ be removed without restarting the node. (Bug #18352)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The cluster created a crashed
+ replica of a table having an ordered index — or when
+ logging was not enabled, of a table having a table or unique
+ index — leading to a crash of the cluster following 8
+ successibe restarts. (Bug #18298)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: When replacing a failed master
+ node, the replacement node could cause the cluster to crash
+ from a buffer overflow if it had an excessively large amount
+ of data to write to the cluster log. (Bug #18118)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Restarting nodes were allowed
+ to start and join the cluster too early. (Bug #16772)
+ </para>
+ </listitem>
<listitem>
<para>
- If InnoDB encountered a
+ If <literal>InnoDB</literal> encountered a
<literal>HA_ERR_LOCK_TABLE_FULL</literal> error and
- rolled-back a transaction, the transaction was still written
+ rolled back a transaction, the transaction was still written
to the binary log. (Bug #18283)
</para>
</listitem>
Modified: trunk/refman-common/news-5.0.xml
===================================================================
--- trunk/refman-common/news-5.0.xml 2006-03-27 13:48:57 UTC (rev 1684)
+++ trunk/refman-common/news-5.0.xml 2006-03-27 14:04:41 UTC (rev 1685)
@@ -181,6 +181,15 @@
</remark>
<itemizedlist>
+
+ <listitem>
+ <para>
+ The <literal>NDBCluster</literal> storage engine now supports
+ <literal>INSERT IGNORE</literal> and <literal>REPLACE</literal>
+ statements. Previously, these statements failed with an error. (Bug
+ #17431)
+ </para>
+ </listitem>
<listitem>
<para>
@@ -224,6 +233,70 @@
</para>
<itemizedlist>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: A timeout in the handling of
+ an <literal>ABORT</literal> condition with more that 32
+ operations could yield a node failure. (Bug #18414)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: A node restart immediately
+ following a <literal>CREATE TABLE</literal> would fail.
+ <emphasis role="bold">Important</emphasis>: This fix supports
+ 2-node Clusters only. (Bug #18385)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: In event of a node failure
+ during a rollback, a <quote>false</quote> lock could be
+ established on the backup for that node, which lock could not
+ be removed without restarting the node. (Bug #18352)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The cluster created a crashed
+ replica of a table having an ordered index — or when
+ logging was not enabled, of a table having a table or unique
+ index — leading to a crash of the cluster following 8
+ successibe restarts. (Bug #18298)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: When replacing a failed master
+ node, the replacement node could cause the cluster to crash
+ from a buffer overflow if it had an excessively large amount
+ of data to write to the cluster log. (Bug #18118)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: If a <command>mysql</command>
+ or other client could not parse the result set returned from a
+ <command>mysqld</command> process acting as an SQL node in a
+ cluster, the client would crash instead of returning the
+ appropriate error. For example, this could happen when the
+ client attempted to use a character set was not available to
+ the <command>mysqld</command>. (Bug #17380)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Restarting nodes were allowed
+ to start and join the cluster too early. (Bug #16772)
+ </para>
+ </listitem>
<listitem>
<para>
Modified: trunk/refman-common/news-5.1.xml
===================================================================
--- trunk/refman-common/news-5.1.xml 2006-03-27 13:48:57 UTC (rev 1684)
+++ trunk/refman-common/news-5.1.xml 2006-03-27 14:04:41 UTC (rev 1685)
@@ -71,13 +71,19 @@
Functionality added or changed:
</para>
-<!--
<itemizedlist>
+
<listitem>
- <para></para>
+ <para>
+ The <literal>NDBCluster</literal> storage engine now supports
+ <literal>CREATE TABLE</literal> statements of arbitrary length.
+ (Previously, <literal>CREATE TABLE</literal> statements for MySQL
+ Cluster tables could contain a maximum of 4096 characters only.) (Bug
+ #17813)
+ </para>
</listitem>
+
</itemizedlist>
--->
<para>
Bugs fixed:
@@ -87,9 +93,9 @@
<listitem>
<para>
- If InnoDB encountered a
+ If <literal>InnoDB</literal> encountered a
<literal>HA_ERR_LOCK_TABLE_FULL</literal> error and
- rolled-back a transaction, the transaction was still written
+ rolled back a transaction, the transaction was still written
to the binary log. (Bug #18283)
</para>
</listitem>
@@ -237,6 +243,15 @@
#17754)
</para>
</listitem>
+
+ <listitem>
+ <para>
+ The <literal>NDBCluster</literal> storage engine now supports
+ <literal>INSERT IGNORE</literal> and <literal>REPLACE</literal>
+ statements. Previously, these statements failed with an error. (Bug
+ #17431)
+ </para>
+ </listitem>
<listitem>
<para>
@@ -476,13 +491,74 @@
<listitem>
<para>
- Using triggers with partitioned InnoDB tables led to incorrect
- results. (Bug #17744)
+ <literal>NDB Cluster</literal>: A timeout in the handling of
+ an <literal>ABORT</literal> condition with more that 32
+ operations could yield a node failure. (Bug #18414)
</para>
</listitem>
<listitem>
<para>
+ <literal>NDB Cluster</literal>: A node restart immediately
+ following a <literal>CREATE TABLE</literal> would fail.
+ <emphasis role="bold">Important</emphasis>: This fix supports
+ 2-node Clusters only. (Bug #18385)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: In event of a node failure
+ during a rollback, a <quote>false</quote> lock could be
+ established on the backup for that node, which lock could not
+ be removed without restarting the node. (Bug #18352)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: The cluster created a crashed
+ replica of a table having an ordered index — or when
+ logging was not enabled, of a table having a table or unique
+ index — leading to a crash of the cluster following 8
+ successibe restarts. (Bug #18298)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: When replacing a failed master
+ node, the replacement node could cause the cluster to crash
+ from a buffer overflow if it had an excessively large amount
+ of data to write to the cluster log. (Bug #18118)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Restarting nodes were allowed to start
+ and join the cluster too early. (Bug #16772)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ <literal>NDB Cluster</literal>: Issuing a <literal>DROP LOGFILE
+ GROUP</literal> statement would cause <command>ndbd</command>
+ processes to crash if MySQL had been compiled with
+ <command>gcc4</command>. (Bug #18295)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Using triggers with partitioned <literal>InnoDB</literal> tables led
+ to incorrect results. (Bug #17744)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
Calling <literal>CREATE TABLE</literal> or <literal>ALTER
TABLE</literal> twice on a partitioned table in a stored
procedure or a prepared statement resulted in errors and
| Thread |
|---|
| • svn commit - mysqldoc@docsrva: r1685 - in trunk: refman-5.0 refman-5.1 refman-common | jon | 27 Mar |
