Author: jstephens
Date: 2006-06-21 06:26:19 +0200 (Wed, 21 Jun 2006)
New Revision: 2449
Log:
Cluster data nodes/limitations - make it clear that DataMemory is
allocated to specific tables and once used cant be freed except by
deleting that table (Hartmut/Tomas)
Modified:
trunk/refman-4.1/ndbcluster.xml
trunk/refman-5.0/ndbcluster.xml
trunk/refman-5.1/ndbcluster.xml
Modified: trunk/refman-4.1/ndbcluster.xml
===================================================================
--- trunk/refman-4.1/ndbcluster.xml 2006-06-20 20:02:24 UTC (rev 2448)
+++ trunk/refman-4.1/ndbcluster.xml 2006-06-21 04:26:19 UTC (rev 2449)
@@ -3551,12 +3551,17 @@
normally partitioned into the same number of fragments as
there are data nodes in the cluster. Thus, for each node,
there are the same number of fragments as are set in
- <literal>NoOfReplicas</literal>. Once a page has been
+ <literal>NoOfReplicas</literal>.
+ </para>
+
+ <para>Once a page has been
allocated, it is currently not possible to return it to
the pool of free pages, except by deleting the table.
- Performing a node recovery also compresses the partition
- because all records are inserted into empty partitions
- from other live nodes.
+ (This also means that <literal>DataMemory</literal> pages,
+ once allocated to a given table, cannot be used by other
+ tables.) Performing a node recovery also compresses the
+ partition because all records are inserted into empty
+ partitions from other live nodes.
</para>
<para>
@@ -12837,6 +12842,19 @@
(<literal>DataMemory</literal> and
<literal>IndexMemory</literal>, respectively).
</para>
+
+ <para>
+ <literal>DataMemory</literal> is allocated as 32KB
+ pages. As each <literal>DataMemory</literal> page is
+ used, it is assigned to a specific table; once
+ allocated, this memory cannot be freed except by
+ deleting the table.
+ </para>
+
+ <para>
+ See <xref linkend="mysql-cluster-db-definition"/>, for
+ further information about <literal>DataMemory</literal> and <literal>IndexMemory</literal>.
+ </para>
</listitem>
<listitem>
Modified: trunk/refman-5.0/ndbcluster.xml
===================================================================
--- trunk/refman-5.0/ndbcluster.xml 2006-06-20 20:02:24 UTC (rev 2448)
+++ trunk/refman-5.0/ndbcluster.xml 2006-06-21 04:26:19 UTC (rev 2449)
@@ -3538,12 +3538,17 @@
normally partitioned into the same number of fragments as
there are data nodes in the cluster. Thus, for each node,
there are the same number of fragments as are set in
- <literal>NoOfReplicas</literal>. Once a page has been
+ <literal>NoOfReplicas</literal>.
+ </para>
+
+ <para>Once a page has been
allocated, it is currently not possible to return it to
the pool of free pages, except by deleting the table.
- Performing a node recovery also compresses the partition
- because all records are inserted into empty partitions
- from other live nodes.
+ (This also means that <literal>DataMemory</literal> pages,
+ once allocated to a given table, cannot be used by other
+ tables.) Performing a node recovery also compresses the
+ partition because all records are inserted into empty
+ partitions from other live nodes.
</para>
<para>
@@ -12889,6 +12894,19 @@
(<literal>DataMemory</literal> and
<literal>IndexMemory</literal>, respectively).
</para>
+
+ <para>
+ <literal>DataMemory</literal> is allocated as 32KB
+ pages. As each <literal>DataMemory</literal> page is
+ used, it is assigned to a specific table; once
+ allocated, this memory cannot be freed except by
+ deleting the table.
+ </para>
+
+ <para>
+ See <xref linkend="mysql-cluster-db-definition"/>, for
+ further information about <literal>DataMemory</literal> and <literal>IndexMemory</literal>.
+ </para>
</listitem>
<listitem>
Modified: trunk/refman-5.1/ndbcluster.xml
===================================================================
--- trunk/refman-5.1/ndbcluster.xml 2006-06-20 20:02:24 UTC (rev 2448)
+++ trunk/refman-5.1/ndbcluster.xml 2006-06-21 04:26:19 UTC (rev 2449)
@@ -3538,12 +3538,17 @@
normally partitioned into the same number of fragments as
there are data nodes in the cluster. Thus, for each node,
there are the same number of fragments as are set in
- <literal>NoOfReplicas</literal>. Once a page has been
+ <literal>NoOfReplicas</literal>.
+ </para>
+
+ <para>Once a page has been
allocated, it is currently not possible to return it to
the pool of free pages, except by deleting the table.
- Performing a node recovery also compresses the partition
- because all records are inserted into empty partitions
- from other live nodes.
+ (This also means that <literal>DataMemory</literal> pages,
+ once allocated to a given table, cannot be used by other
+ tables.) Performing a node recovery also compresses the
+ partition because all records are inserted into empty
+ partitions from other live nodes.
</para>
<para>
@@ -15305,6 +15310,19 @@
(<literal>DataMemory</literal> and
<literal>IndexMemory</literal>, respectively).
</para>
+
+ <para>
+ <literal>DataMemory</literal> is allocated as 32KB
+ pages. As each <literal>DataMemory</literal> page is
+ used, it is assigned to a specific table; once
+ allocated, this memory cannot be freed except by
+ deleting the table.
+ </para>
+
+ <para>
+ See <xref linkend="mysql-cluster-db-definition"/>, for
+ further information about <literal>DataMemory</literal> and <literal>IndexMemory</literal>.
+ </para>
</listitem>
<listitem>
| Thread |
|---|
| • svn commit - mysqldoc@docsrva: r2449 - in trunk: refman-4.1 refman-5.0 refman-5.1 | jon | 21 Jun |
