安装配置Greenplum

一、规划

1、服务器：

192.168.0.191   gpdb01        # master

192.168.0.192   gpdb02        # primary segment 、 mirror segment

192.168.0.193   gpdb03        # primary segment 、 mirror segment

192.168.0.194   gpdb04        # primary segment 、 mirror segment

192.168.0.195   gpdb05        # standy

2、安装基本服务

yum -y install vim ed

二、配置

1、关闭防火墙

sed -i '/SELINUX=enforcing/d' /etc/selinux/config

sed -i '/SELINUX=disabled/d' /etc/selinux/config

echo "SELINUX=disabled" >> /etc/selinux/config

sed -e 's/SELINUX=enforcing/SELINUX=disabled/d' /etc/selinux/config

systemctl stop firewalld

chkconfig firewalld off

2、配置/etc/hosts

192.168.0.191   gpdb01

192.168.0.192   gpdb02

192.168.0.193   gpdb03

192.168.0.194   gpdb04

192.168.0.195   gpdb05

3、创建用户

groupadd gpadmin

useradd -g gpadmin gpadmin

echo "gpadmin" | passwd --stdin gpadmin

id gpadmin

三、以root用户在master节点安装greenplum

1、准备安装文件

greenplum-db-4.3.5.1-build-1-RHEL5-x86_64.bin

2、创建安装目录

mkdir -p /opt/greenplum

chown -R gpadmin:gpadmin /opt/greenplum/

3、安装greenplum

./greenplum-db-4.3.5.1-build-1-RHEL5-x86_64.bin

... ...

... ...

********************************************************************************

Do you accept the Pivotal Database license agreement? [yes|no]

********************************************************************************

yes

********************************************************************************

Provide the installation path for Greenplum Database or press ENTER to

accept the default installation path: /usr/local/greenplum-db-4.3.5.1

********************************************************************************

/opt/greenplum/greenplum-db-4.3.5.1

********************************************************************************

Install Greenplum Database into </opt/greenplum/greenplum-db-4.3.5.1>? [yes|no]

********************************************************************************

yes

********************************************************************************

/opt/greenplum/greenplum-db-4.3.5.1 does not exist.

Create /opt/greenplum/greenplum-db-4.3.5.1 ? [yes|no]

(Selecting no will exit the installer)

********************************************************************************

yes

Extracting product to /opt/greenplum/greenplum-db-4.3.5.1

********************************************************************************

Installation complete.

Greenplum Database is installed in /opt/greenplum/greenplum-db-4.3.5.1

Pivotal Greenplum documentation is available

for download at http://docs.gopivotal.com/gpdb

********************************************************************************

安装目录如下：

[root@gpdb01 greenplum]# pwd

/opt/greenplum

[root@gpdb01 greenplum]# ll

total 0

lrwxrwxrwx.  1 root root  22 Mar  2 05:34 greenplum-db -> ./greenplum-db-4.3.5.1

drwxr-xr-x. 11 root root 189 Mar  2 05:34 greenplum-db-4.3.5.1

[root@gpdb01 greenplum]# ll greenplum-db/

total 264

drwxr-xr-x. 3 gpadmin gpadmin   4096 May 15  2015 bin

drwxr-xr-x. 2 gpadmin gpadmin     64 May 15  2015 demo

drwxr-xr-x. 5 gpadmin gpadmin     52 May 15  2015 docs

drwxr-xr-x. 2 gpadmin gpadmin     44 May 15  2015 etc

drwxr-xr-x. 3 gpadmin gpadmin     20 May 15  2015 ext

-rw-r--r--. 1 gpadmin gpadmin  43025 May 15  2015 GPDB-LICENSE.txt

-rw-r--r--. 1 root    root       680 Mar  2 05:34 greenplum_path.sh

drwxr-xr-x. 6 gpadmin gpadmin   4096 May 15  2015 include

drwxr-xr-x. 9 gpadmin gpadmin   8192 May 15  2015 lib

-rw-r--r--. 1 gpadmin gpadmin 193083 May 15  2015 LICENSE.thirdparty

drwxr-xr-x. 2 gpadmin gpadmin   4096 May 15  2015 sbin

drwxr-xr-x. 3 gpadmin gpadmin     24 May 15  2015 share

[root@gpdb01 greenplum]#

4、应用greenplum的环境变量

source /opt/greenplum/greenplum-db/greenplum_path.sh

5、配置hostlist和segmentlist

cd /opt/greenplum/greenplum-db

mkdir conf

cat conf/hostlist

gpdb01

gpdb02

gpdb03

gpdb04

gpdb05

cat conf/segmentlists

gpdb02

gpdb03

gpdb04

6、使用gpssh-exkeys打通所有服务器

gpssh-exkeys -f ./conf/hostlist

[STEP 1 of 5] create local ID and authorize on local host

[STEP 2 of 5] keyscan all hosts and update known_hosts file

[STEP 3 of 5] authorize current user on remote hosts
   ... send to gpdb02
   ***
   *** Enter password for gpdb02:
   ... send to gpdb03
   ... send to gpdb04
   ... send to gpdb05

[STEP 4 of 5] determine common authentication file content

[STEP 5 of 5] copy authentication files to all remote hosts
   ... finished key exchange with gpdb02
   ... finished key exchange with gpdb03
   ... finished key exchange with gpdb04
   ... finished key exchange with gpdb05

[INFO] completed successfully

打通服务器后可以使用gpssh进行批量操作了：

gpssh -f ./conf/hostlist

Note: command history unsupported on this machine ...

=> pwd

[gpdb05] /home/gpadmin

[gpdb04] /home/gpadmin

[gpdb01] /home/gpadmin

[gpdb03] /home/gpadmin

[gpdb02] /home/gpadmin

=>

7、将安装后的软件分发的每一台服务器上并创建相应目录：

tar -cf greenplum-db-4.3.5.1.tar ./greenplum-db-4.3.5.1

gpscp -f /opt/greenplum/greenplum-db/conf/hostlist ../greenplum-db-4.3.5.1.tar =:/opt/greenplum

gpssh -f /opt/greenplum/greenplum-db/conf/hostlist

Note: command history unsupported on this machine ...

=> cd /opt/greenplum

[gpdb05]

[gpdb04]

[gpdb01]

[gpdb03]

[gpdb02]

=> tar -xf greenplum-db-4.3.5.1.tar

[gpdb05]

[gpdb04]

[gpdb01]

[gpdb03]

[gpdb02]

=> ln -s /opt/greenplum/greenplum-db-4.3.5.1 /opt/greenplum/greenplum-db

[gpdb05]

[gpdb04]

[gpdb01]

[gpdb03]

[gpdb02]

=> mkdir -p /home/gpadmin/gpdata/gpmaster

[gpdb05]

[gpdb04]

[gpdb01]

[gpdb03]

[gpdb02]

=> mkdir -p /home/gpadmin/gpdata/gpdatap1

[gpdb05]

[gpdb04]

[gpdb01]

[gpdb03]

[gpdb02]

=> mkdir -p /home/gpadmin/gpdata/gpdatam1

[gpdb05]

[gpdb04]

[gpdb01]

[gpdb03]

[gpdb02]

=>

8、配置环境变量：

vim ~/.bash_profile

source /opt/greenplum/greenplum-db/greenplum_path.sh

export MASTER_DATA_DIRECTORY=/home/gpadmin/gpdata/gpmaster/gpseg-1

export PGPORT=2345

export PGDATABASE=chavin

9、初始化greenplum配置文件

cp $GPHOME/docs/cli_help/gpconfigs/gpinitsystem_config ./conf/gpinitsystem_config

cat conf/gpinitsystem_config

ARRAY_NAME="Greenplum"

SEG_PREFIX=gpseg

PORT_BASE=40000

declare -a DATA_DIRECTORY=(/home/gpadmin/gpdata/gpdatap1)

MASTER_HOSTNAME=gpdb01

MASTER_DIRECTORY=/home/gpadmin/gpdata/gpmaster

MASTER_PORT=2345

TRUSTED_SHELL=ssh

CHECK_POINT_SEGMENTS=8

ENCODING=UNICODE

MIRROR_PORT_BASE=50000

REPLICATION_PORT_BASE=41000

MIRROR_REPLICATION_PORT_BASE=51000

declare -a MIRROR_DATA_DIRECTORY=(/home/gpadmin/gpdata/gpdatam1)

#DATABASE_NAME=name_of_database

MACHINE_LIST_FILE=/opt/greenplum/greenplum-db/conf/segmentlists

10、初始化数据库

gpinitsystem -c ./conf/gpinitsystem_config -s gpdb05

gpstate -s

$ psql -d postgres

psql (8.2.15)

Type "help" for help.

postgres=# create database chavin;

CREATE DATABASE

-- 测试表

drop table if exists dept;

create table dept

(

deptno int primary key,

dname varchar(14),

loc varchar(13)

);

insert into dept(deptno,dname,loc) values(10,'ACCOUNTING','NEW YORK');

insert into dept(deptno,dname,loc) values(20,'RESEARCH','DALLAS');

insert into dept(deptno,dname,loc) values(30,'SALES','CHICAGO');

insert into dept(deptno,dname,loc) values(40,'OPERATIONS','BOSTON');

$ psql

psql (8.2.15)

Type "help" for help.

chavin=# create table dept

chavin-# (

chavin(# deptno int primary key,

chavin(# dname varchar(14),

chavin(# loc varchar(13)

chavin(# );

NOTICE:  CREATE TABLE / PRIMARY KEY will create implicit index "dept_pkey" for table "dept"

CREATE TABLE

chavin=# insert into dept(deptno,dname,loc) values(10,'ACCOUNTING','NEW YORK');

INSERT 0 1

chavin=# insert into dept(deptno,dname,loc) values(20,'RESEARCH','DALLAS');

INSERT 0 1

chavin=# insert into dept(deptno,dname,loc) values(30,'SALES','CHICAGO');

INSERT 0 1

chavin=# insert into dept(deptno,dname,loc) values(40,'OPERATIONS','BOSTON');

INSERT 0 1

chavin=#

启动关闭数据库：

gpstart -a

gpstop -a

出现的问题：

1、如果第一次初始化失败，再次初始化报如下错误：

20180302:08:00:01:002185 gpinitsystem:gpdb01:gpadmin-[WARN]:-Have lock file /tmp/.s.PGSQL.2345.lock but no process running on port 2345

20180302:08:00:01:gpinitsystem:gpdb01:gpadmin-[FATAL]:-Found indication of postmaster process on port 2345 on Master host Script Exiting!

删除/tmp目录下的隐含文件/tmp/.s.PGSQL.2345.lock后，重新初始化即可。

2、重新初始化出现如下错误：

20180302:18:49:37:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Waiting for parallel processes batch [1], please wait...

.....................................................................................

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:------------------------------------------------

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Parallel process exit status

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:------------------------------------------------

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Total processes marked as completed           = 0

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Total processes marked as killed              = 0

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[WARN]:-Total processes marked as failed              = 6 <<<<<

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:------------------------------------------------

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Commencing parallel build of mirror segment instances

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Spawning parallel processes    batch [1], please wait...

......

20180302:18:51:03:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Waiting for parallel processes batch [1], please wait...

........

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:------------------------------------------------

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Parallel process exit status

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:------------------------------------------------

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Total processes marked as completed           = 0

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Total processes marked as killed              = 0

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[WARN]:-Total processes marked as failed              = 6 <<<<<

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:------------------------------------------------

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[FATAL]:-Errors generated from parallel processes

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Dumped contents of status file to the log file

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Building composite backout file

20180302:18:51:12:gpinitsystem:gpdb01:gpadmin-[FATAL]:-Failures detected, see log file /home/gpadmin/gpAdminLogs/gpinitsystem_20180302.log for more detail Script Exiting!

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[WARN]:-Script has left Greenplum Database in an incomplete state

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[WARN]:-Run command /bin/bash /home/gpadmin/gpAdminLogs/backout_gpinitsystem_gpadmin_20180302_184756 to remove these changes

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-Start Function BACKOUT_COMMAND

20180302:18:51:12:003425 gpinitsystem:gpdb01:gpadmin-[INFO]:-End Function BACKOUT_COMMAND

这是因为缺少ed依赖包，通过yum -y install ed后重启服务器，可以解决。

3、gpstart启动报错如下：

20180302:20:33:28:003078 gpstart:gpdb01:gpadmin-[ERROR]:-gpstart error: Do not have enough valid segments to start the array.

这个时候需要检查一下shared_buffers设置改小点，就行了