其他见：http://my.oschina.net/ensn/blog/636766
本文利用pg_upgrade实现将8.4.18版本升级到9.5.0版本，8.4.18版本为RedHat系统自带pg数据库版本。
环境：Red Hat Enterprise Linux Server release 6.5 (Santiago) X86-64
查看手册看到可以利用pg_upgrade实现从8.4到9.5版本直接升级：
pg_upgrade supports upgrades from 8.3.X and later to the current major release of PostgreSQL, including snapshot and alpha releases.
下载postgresql9.5.0版本：
1.解压：
tar jxvf postgresql-9.5.0.tar.bz2
2.编译：
./configure --prefix=/usr/local/pgsql9.5.0 --with-pgport=5434 --with-wal-blocksize=16 //这里在编译时带with-wal-blocksize=16配置参数导致后面升级出现错误，具体到后面解释。注意指定目录，方便后面管理
***编译过程可能碰到错误：
configure: error: readline library notfound
If you have readline already installed, see config.log for detailson the
failure. It is possible the compiler isnt lookingin the proper directory.
Use --without-readline to disable readlinesupport.
解决方法：
rpm -qa | grep readline
yum search readline
找到符合配置的版本按照：
yum -y install -y readline-deve
安装完毕readline，重新执行上面编译命令。
3.安装：
gmake world
gmake install -world //此处要加world
4.初始化数据库：
4.1 创建data数据目录：（可另指定其他位置，此处将指定在pg安装目录）
cd /usr/local/pgsql9.5.0
mkdir data
4.2 修改data目录权限，将其赋给postgres：
[root@localhost pgsql9.5.0]# chown -R postgres:postgres /usr/local/pgsql9.5.0/data
[root@localhost pgsql9.5.0]# chmod -R 700 /usr/local/pgsql9.5.0/data
4.3 初始化：
-bash-4.1$ /usr/local/pgsql9.5.0/bin/initdb -E UTF8 -D /usr/local/pgsql9.5.0/data --locale=C -U postgres -W //locale=C这里在升级时报错，后续具体解释
The files belonging to this database system will be owned by user "postgres".
This user must also own the server process.
The database cluster will be initialized with locale "C".
The default text search configuration will be set to "english".
Data page checksums are disabled.
....
初始化过程中会要求设置一个新的密码：postgres
...
Success. You can now start the database server using:
/usr/local/pgsql9.5.0/bin/pg_ctl -D /usr/local/pgsql9.5.0/data -l logfile start
***********************
：/usr/local/pgsql9.5.0/bin/为安装目录，/usr/local/pgsql9.5.0/data为数据目录
:初始化需要在postgres用户下进行，不能再root下
[root@localhost pgsql9.5.0]# /usr/local/pgsql9.5.0/bin/initdb -E UTF8 -D /usr/local/pgsql9.5.0/data --locale=C -U postgres -W
initdb: cannot be run as root
Please log in (using, e.g., "su") as the (unprivileged) user that will
own the server process.
************************
5.安装pg_upgrade插件：
cd contrib
gamke
gmake install
5.修改postgresql.conf pg_hba.conf文件：
因为升级需要多次连接新老集群数据库实例, 所以修改为使用本地trust认证.
host all all 127.0.0.1/32 trust
端口使用注意使用不同的监听端口.（使用编译时候的指定端口：5434）
6.停老库：(新库没有启动，故这里不停)
-bash-4.1$ pg_ctl stop -m fast -D /var/lib/pgsql/data
waiting for server to shut down.... done
server stopped
7.更新：
首先创建一个用户执行升级的目录，该目录权限给postgres用户（执行更新时，会产一些升级文件，创建upgrade目录存储该些文件）
su - root
cd /usr/local/pgsql9.5.0
mkdir upgrade
chown postgres：postgres upgrade
su - postgres
cd /usr/local/pgsql9.5.0/upgrade
7.1检验更新：
-bash-4.1$ /usr/local/pgsql9.5.0/bin/pg_upgrade -c -b /usr/bin -B /usr/local/pgsql9.5.0/bin -d /var/lib/pgsql/data -D /usr/local/pgsql9.5.0/data -p 5432 -P 5434 -U postgres //-c表示检验，-b/-B 表示老/新库bin目录，-d/-D表示老/新库数据目录，具体命令参考手册
检验时碰见以下问题：
1).wal-blocksize 问题
该问题是由前后两个版本wal-blocksize不同，因为之前在安装新版本时指定了wal-blocksize的大小为16k，但老版本是默认安装的，默认大小为8k，该参数大小只能在安装时改变，故需要重新安装pg数据库来解决该问题。
2).lc_collate values for database "postgres" do not match: old "en_US.UTF-8", new "C" Failure, exiting
删除data数据目录，重新做一次初始化
/usr/local/pgsql9.5.0/bin/initdb -E UTF8 -D /usr/local/pgsql9.5.0/data --locale=en_US.UTF-8 -U postgres -W
解决以上问题以后重新检验更新:
-bash-4.1$ /usr/local/pgsql9.5.0/bin/pg_upgrade -c -b /usr/bin -B /usr/local/pgsql9.5.0/bin -d /var/lib/pgsql/data -D /usr/local/pgsql9.5.0/data -p 5432 -P 5434 -U postgres
Performing Consistency Checks
-----------------------------
Checking cluster versions ok
Checking database user is the install user ok
Checking database connection settings ok
Checking for prepared transactions ok
Checking for reg* system OID user data types ok
Checking for contrib/isn with bigint-passing mismatch ok
Checking for invalid "line" user columns ok
Checking for large objects ok
Checking for presence of required libraries ok
Checking database user is the install user ok
Checking for prepared transactions ok
*Clusters are compatible*
以上说明可以执行更新。
7.2 执行更新
-bash-4.1$ /usr/local/pgsql9.5.0/bin/pg_upgrade -b /usr/bin -B /usr/local/pgsql9.5.0/bin -d /var/lib/pgsql/data -D /usr/local/pgsql9.5.0/data -p 5432 -P 5434
*****
pg_upgrade有两种升级方式，一个是缺省的通过拷贝数据文件到新的data目录下，一个是创建硬链接。拷贝的方式升级较慢，但是原库还可用；硬链接的方式升级较快，但是原库不可用。以上是缺省的方式，硬链接方式升级的命令只需要添加--link
Performing Consistency Checks
-----------------------------
Checking cluster versions ok
Checking database user is the install user ok
Checking database connection settings ok
Checking for prepared transactions ok
Checking for reg* system OID user data types ok
Checking for contrib/isn with bigint-passing mismatch ok
Checking for invalid "line" user columns ok
Checking for large objects ok
Creating dump of global objects ok
Creating dump of database schemas
ok
Checking for presence of required libraries ok
Checking database user is the install user ok
Checking for prepared transactions ok
If pg_upgrade fails after this point, you must re-initdb the
new cluster before continuing.
Performing Upgrade
------------------
Analyzing all rows in the new cluster ok
Freezing all rows on the new cluster ok
Deleting files from new pg_clog ok
Copying old pg_clog to new server ok
Setting next transaction ID and epoch for new cluster ok
Deleting files from new pg_multixact/offsets ok
Setting oldest multixact ID on new cluster ok
Resetting WAL archives ok
Setting frozenxid and minmxid counters in new cluster ok
Restoring global objects in the new cluster ok
Restoring database schemas in the new cluster
ok
Setting minmxid counter in new cluster ok
Creating newly-required TOAST tables ok
Copying user relation files
ok
Setting next OID for new cluster ok
Sync data directory to disk ok
Creating script to analyze new cluster ok
Creating script to delete old cluster ok
Checking for large objects ok
Upgrade Complete
----------------
Optimizer statistics are not transferred by pg_upgrade so,
once you start the new server, consider running:
./analyze_new_cluster.sh
Running this script will delete the old cluster's data files:
./delete_old_cluster.sh
升级过程是拷贝原来目录的数据到新版本指定的目录中，最后提示生成两个脚本，一个是analyze_new_cluster.sh，需要在新版本中执行，用来收集统计信息，另一个是 delete_old_cluster.sh，用来删除旧版本集群数据，当然为了安全起见可以等系统运行几天没问题再来删除。
8.启动新数据库
-bash-4.1$ pg_ctl start -D /usr/local/pgsql9.5.0/data
报错：
server starting
-bash-4.1$ FATAL: unrecognized configuration parameter "dynamic_shared_memory_type"
解决方法：
原因是因为版本不一致导致，可能是我的主机上面被我编译了两个版本导致的，之前８.４postgresql.conf文件中没有该参数，升级后9.5的配置文件中有该参数，故出现不一致，
通过绝对路径启动数据库：
[root@localhost ~]# su - postgres
-bash-4.1$ /usr/local/pgsql9.5.0/bin/pg_ctl start -D /usr/local/pgsql9.5.0/data
server starting
-bash-4.1$ LOG: database system was shut down at 2016-03-08 10:55:18 CST
LOG: MultiXact member wraparound protections are now enabled
LOG: autovacuum launcher started
LOG: database system is ready to accept connections
进入数据库，此时psql无法进入：
-bash-4.1$ psql
psql: could not connect to server: No such file or directory
Is the server running locally and accepting
connections on Unix domain socket "/tmp/.s.PGSQL.5432"?
通过指定psql路径进入：
-bash-4.1$ /usr/local/pgsql9.5.0/bin/psql --指定路径
psql (9.5.0)
Type "help" for help.
postgres=#
9.执行收集统计信息的脚本：
-bash-4.1$ /usr/local/pgsql9.5.0/upgrade/analyze_new_cluster.sh
This script will generate minimal optimizer statistics rapidly
so your system is usable, and then gather statistics twice more
with increasing accuracy. When it is done, your system will
have the default level of optimizer statistics.
If you have used ALTER TABLE to modify the statistics target for
any tables, you might want to remove them and restore them after
running this script because they will delay fast statistics generation.
If you would like default statistics as quickly as possible, cancel
this script and run:
"/usr/local/pgsql9.5.0/bin/vacuumdb" --all --analyze-only
vacuumdb: processing database "gedbs": Generating minimal optimizer statistics (1 target)
vacuumdb: processing database "postgres": Generating minimal optimizer statistics (1 target)
vacuumdb: processing database "template1": Generating minimal optimizer statistics (1 target)
vacuumdb: processing database "gedbs": Generating medium optimizer statistics (10 targets)
vacuumdb: processing database "postgres": Generating medium optimizer statistics (10 targets)
vacuumdb: processing database "template1": Generating medium optimizer statistics (10 targets)
vacuumdb: processing database "gedbs": Generating default (full) optimizer statistics
vacuumdb: processing database "postgres": Generating default (full) optimizer statistics
vacuumdb: processing database "template1": Generating default (full) optimizer statistics
Done
10.修改bash_profile文件中的PGDATA， PGPORT等参数（对postgres用户有效，客户端用其他用户登陆可能无效，修改/etc/profile对所有用户有效），修改PATH搜索路径，
不然在执行psql是会报版本不一致警告：
WARNING: psql version 8.4, server version 9.5.
11.启动数据库，pgadmin连接
报错：configure FATAL: password authentication failed for user "postgres"
解决方法：
修改postgres用户密码，ALTER USER postgres PASSWORD 'newPassword';
参考：http://*.com/questions/7695962/postgresql-password-authentication-failed-for-user-postgres
12.新数据库稳定后，删除老版本软件
[root@localhost upgrade]# cat delete_old_cluster.sh
#!/bin/sh
rm -rf '/var/lib/pgsql/data'
$ ./delete_old_cluster.sh
参考：
http://www.postgresql.org/docs/current/static/pgupgrade.html
http://my.oschina.net/liuyuanyuangogo/blog/500229
http://xmarker.blog.163.com/blog/static/22648405720145203428811/
http://906179271.iteye.com/blog/2263218
http://francs3.blog.163.com/blog/static/40576727201519521050/
http://wangwei.info/upgrade-postgresql-centos/