Wed Nov 15 14:07:01 PST 2006
- Previous message: [Slony1-commit] By cbbrowne: Further portability changes - Solaris needs to use
- Next message: [Slony1-commit] By cbbrowne: More notes to explain the implications of upgrading from
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Log Message:
-----------
More notes explaining the implications of upgrading from one version of
Slony-I to another
Modified Files:
--------------
slony1-engine/doc/adminguide:
slonyupgrade.sgml (r1.3 -> r1.4)
-------------- next part --------------
Index: slonyupgrade.sgml
===================================================================
RCS file: /usr/local/cvsroot/slony1/slony1-engine/doc/adminguide/slonyupgrade.sgml,v
retrieving revision 1.3
retrieving revision 1.4
diff -Ldoc/adminguide/slonyupgrade.sgml -Ldoc/adminguide/slonyupgrade.sgml -u -w -r1.3 -r1.4
--- doc/adminguide/slonyupgrade.sgml
+++ doc/adminguide/slonyupgrade.sgml
@@ -23,6 +23,14 @@
<listitem><para> Start all slons. </para> </listitem>
</itemizedlist>
+<para> The overall operation is relatively safe: If there is any
+mismatch between component versions, the &lslon; will refuse to start
+up, which provides protection against corruption. </para>
+
+<para> You need to be sure that the C library containing SPI trigger
+functions has been copied into place in the &postgres; build. There
+are multiple possible approaches to this:</para>
+
<para> The trickiest part of this is ensuring that the C library
containing SPI functions is copied into place in the &postgres; build;
the easiest and safest way to handle this is to have two separate
@@ -37,6 +45,35 @@
work on <trademark>Windows</trademark> if it locks library files that
are in use.</para>
+<variablelist>
+
+<varlistentry><term>Run <command>make install</command> to install new
+&slony1; components on top of the old</term>
+
+<listitem><para>If you build &slony1; on the same system on which it
+is to be deployed, and build from sources, overwriting the old with
+the new is as easy as <command>make install</command>. There is no
+need to restart a database backend; just to stop &lslon; processes,
+run the <command>UPDATE FUNCTIONS</command> script, and start new
+&lslon; processes.</para>
+
+<para> Unfortunately, this approach requires having a build
+environment on the same host as the deployment. That may not be
+consistent with efforts to use common &postgres; and &slony1; binaries
+across a set of nodes. </para>
+</listitem></varlistentry>
+
+<varlistentry><term>Create a new &postgres; and &slony1; build</term>
+
+<listitem><para>With this approach, the old &postgres; build with old
+&slony1; components persists after switching to a new &postgres; build
+with new &slony1; components. In order to switch to the new &slony1;
+build, you need to restart the
+&postgres; <command>postmaster</command>, therefore interrupting
+applications, in order to get it to be aware of the location of the
+new components. </para></listitem></varlistentry>
+
+</variablelist>
</sect1>
<!-- Keep this comment at the end of the file
Local variables:
- Previous message: [Slony1-commit] By cbbrowne: Further portability changes - Solaris needs to use
- Next message: [Slony1-commit] By cbbrowne: More notes to explain the implications of upgrading from
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Slony1-commit mailing list