mirrors/postgres
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

New async/sync multi-master headings for docs.

Browse Source
This commit is contained in:
Bruce Momjian 2006-11-22 04:00:19 +00:00
parent ffa0e8f0cd
commit 3b0313580e
1 changed files with 15 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
doc/src/sgml/high-availability.sgml
View File

@ -1,4 +1,4 @@
<!-- $PostgreSQL: pgsql/doc/src/sgml/high-availability.sgml,v 1.9 2006/11/22 03:44:52 momjian Exp $ --> <!-- $PostgreSQL: pgsql/doc/src/sgml/high-availability.sgml,v 1.10 2006/11/22 04:00:19 momjian Exp $ -->
<chapter id="high-availability"> <chapter id="high-availability">
<title>High Availability and Load Balancing</title> <title>High Availability and Load Balancing</title>
@ -184,22 +184,22 @@ protocol to make nodes agree on a serializable transactional order.
</varlistentry> </varlistentry>
<varlistentry> <varlistentry>
<term>Multi-Master Clustering</term> <term>Multi-Master Replication</term>
<listitem> <listitem>
<para> <para>
In clustering, each server can accept write requests, and In multi-master replication, each server can accept write
modified data is transmitted from the original server to every requests, and modified data is transmitted from the original
other server before each transaction commits. Heavy write server to every other server before each transaction commits.
activity can cause excessive locking, leading to poor performance. Heavy write activity can cause excessive locking, leading to
In fact, write performance is often worse than that of a single poor performance. In fact, write performance is often worse
server. Read requests can be sent to any server. Some than that of a single server. Read requests can be sent to
implementations use cluster-wide shared memory or shared disk any server. Some implementations use cluster-wide shared memory
to reduce the communication overhead. Clustering is best for or shared disk to reduce the communication overhead. Clustering
mostly read workloads, though its big advantage is that any is best for mostly read workloads, though its big advantage is
server can accept write requests &mdash; there is no need to that any server can accept write requests &mdash; there is no
partition workloads between master and slave servers, and need to partition workloads between master and slave servers,
because the data changes are sent from one server to another, and because the data changes are sent from one server to another,
there is no problem with non-deterministic functions like there is no problem with non-deterministic functions like
<function>random()</>. <function>random()</>.
</para> </para>
@ -246,7 +246,7 @@ protocol to make nodes agree on a serializable transactional order.
</varlistentry> </varlistentry>
<varlistentry> <varlistentry>
<term>Clustering For Parallel Query Execution</term> <term>Multi-Server Parallel Query Execution</term>
<listitem> <listitem>
<para> <para>