原文链接：https://yq.aliyun.com/articles/695864

[TOC]

概述

新安装的greenplum集群只有primary节点，没有mirror。高可用性没得到保证。所以就需要为集群添加mirror节点

注意：数据量过大时添加mirror，会有较大的磁盘压力，因为后台会一直同步数据，这一点线上系统需要注意

添加前情况

初始化的时候只有primary

gpstate -b

  1.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:-Greenplum Primary Segment Configuration
  2.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:----------------------------------------
  3.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:-gw_sdw1 /data/primary/gpseg0 40000 2 0
  4.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:-gw_sdw1 /data/primary/gpseg1 40001 3 1
  5.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:-gw_sdw1 /data/primary/gpseg2 40002 4 2
  6.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:-gw_sdw1 /data/primary/gpseg3 40003 5 3
  7.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:-gw_sdw2 /data/primary/gpseg4 40000 6 4
  8.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:-gw_sdw2 /data/primary/gpseg5 40001 7 5
  9.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:-gw_sdw2 /data/primary/gpseg6 40002 8 6
  10.  
    20190326:00:48:33:005423 gpinitsystem:gw_mdw1:gpadmin-[INFO]:-gw_sdw2 /data/primary/gpseg7 40003 9 7
  11.  
    Continue with Greenplum creation Yy/Nn>

查看系统配置表，发现只有primary节点

  1.  
    postgres=# SELECT * from gp_segment_configuration ;
  2.  
    dbid | content | role | preferred_role | mode | status | port | hostname | address | replication_port | san_mounts
  3.  
    ------+---------+------+----------------+------+--------+-------+----------+---------+------------------+------------
  4.  
    1 | -1 | p | p | s | u | 5432 | gw_mdw1 | gw_mdw1 | |
  5.  
    2 | 0 | p | p | s | u | 40000 | gw_sdw1 | gw_sdw1 | |
  6.  
    6 | 4 | p | p | s | u | 40000 | gw_sdw2 | gw_sdw2 | |
  7.  
    3 | 1 | p | p | s | u | 40001 | gw_sdw1 | gw_sdw1 | |
  8.  
    7 | 5 | p | p | s | u | 40001 | gw_sdw2 | gw_sdw2 | |
  9.  
    4 | 2 | p | p | s | u | 40002 | gw_sdw1 | gw_sdw1 | |
  10.  
    8 | 6 | p | p | s | u | 40002 | gw_sdw2 | gw_sdw2 | |
  11.  
    5 | 3 | p | p | s | u | 40003 | gw_sdw1 | gw_sdw1 | |
  12.  
    9 | 7 | p | p | s | u | 40003 | gw_sdw2 | gw_sdw2 | |
  13.  
    (9 rows)

添加步骤

在所有需要添加mirror的主机，创建存放mirror的数据目录

可以使用gpssh命令

gpssh -f seg_hosts -e 'mkdir -p /data/mirror'
 

生成addmirror文件

执行gpaddmirrors 命令，键入存放mirror节点的日志，生成配置文件

  1.  
    [gpadmin@gw_mdw1 ~]$ gpaddmirrors -o ./addmirror
  2.  
    20190326:00:56:21:030831 gpaddmirrors:gw_mdw1:gpadmin-[INFO]:-Starting gpaddmirrors with args: -o ./addmirror
  3.  
    20190326:00:56:21:030831 gpaddmirrors:gw_mdw1:gpadmin-[INFO]:-local Greenplum Version: 'postgres (Greenplum Database) 4.3.1.0 build 6'
  4.  
    20190326:00:56:21:030831 gpaddmirrors:gw_mdw1:gpadmin-[INFO]:-master Greenplum Version: 'PostgreSQL 8.2.15 (Greenplum Database 4.3.1.0 build 6) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Jun 11 2014 17:23:40'
  5.  
    20190326:00:56:21:030831 gpaddmirrors:gw_mdw1:gpadmin-[INFO]:-Obtaining Segment details from master...
  6.  
    Enter mirror segment data directory location 1 of 4 >
  7.  
    /data/mirror
  8.  
    Enter mirror segment data directory location 2 of 4 >
  9.  
    /data/mirror
  10.  
    Enter mirror segment data directory location 3 of 4 >
  11.  
    /data/mirror
  12.  
    Enter mirror segment data directory location 4 of 4 >
  13.  
    /data/mirror
  14.  
    20190326:00:57:15:030831 gpaddmirrors:gw_mdw1:gpadmin-[INFO]:-Configuration file output to ./addmirror successfully.

查看文件内容如下

  1.  
    [gpadmin@gw_mdw1 ~]$ cat addmirror
  2.  
    filespaceOrder=
  3.  
    mirror0=0:gw_sdw2:41000:42000:43000:/data/mirror/gpseg0
  4.  
    mirror1=1:gw_sdw2:41001:42001:43001:/data/mirror/gpseg1
  5.  
    mirror2=2:gw_sdw2:41002:42002:43002:/data/mirror/gpseg2
  6.  
    mirror3=3:gw_sdw2:41003:42003:43003:/data/mirror/gpseg3
  7.  
    mirror4=4:gw_sdw1:41000:42000:43000:/data/mirror/gpseg4
  8.  
    mirror5=5:gw_sdw1:41001:42001:43001:/data/mirror/gpseg5
  9.  
    mirror6=6:gw_sdw1:41002:42002:43002:/data/mirror/gpseg6
  10.  
    mirror7=7:gw_sdw1:41003:42003:43003:/data/mirror/gpseg7

执行添加命令

  1.  
    [gpadmin@gw_mdw1 ~]$ gpaddmirrors -i addmirror
  2.  
    20190326:01:08:45:031106 gpaddmirrors:gw_mdw1:gpadmin-[INFO]:-Starting gpaddmirrors with args: -i addmirror
  3.  
    20190326:01:08:45:031106 gpaddmirrors:gw_mdw1:gpadmin-[INFO]:-local Greenplum Version: 'postgres (Greenplum Database) 4.3.1.0 build 6'
  4.  
    20190326:01:08:45:031106 gpaddmirrors:gw_mdw1:gpadmin-[INFO]:-master Greenplum Version: 'PostgreSQL 8.2.15 (Greenplum Database 4.3.1.0 build 6) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Jun 11 2014 17:23:40'

命令没有报错，查看mirror节点的情况

使用gpstate -m查看，发现所有的mirror正在同步数据，因为我的集群新建，很快就同步完成了，此时再执行gpstate -m就可以看到Data Status的状态是Synchronized（已同步的）

  1.  
    [gpadmin@gw_mdw1 ~]$ gpstate -m
  2.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:-Starting gpstate with args: -m
  3.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:-local Greenplum Version: 'postgres (Greenplum Database) 4.3.1.0 build 6'
  4.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:-master Greenplum Version: 'PostgreSQL 8.2.15 (Greenplum Database 4.3.1.0 build 6) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Jun 11 2014 17:23:40'
  5.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:-Obtaining Segment details from master...
  6.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:--------------------------------------------------------------
  7.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:--Current GPDB mirror list and status
  8.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:--Type = Group
  9.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:--------------------------------------------------------------
  10.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:- Mirror Datadir Port Status Data Status
  11.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:- gw_sdw2 /data/mirror/gpseg0 41000 Passive Resynchronizing
  12.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:- gw_sdw2 /data/mirror/gpseg1 41001 Passive Resynchronizing
  13.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:- gw_sdw2 /data/mirror/gpseg2 41002 Passive Resynchronizing
  14.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:- gw_sdw2 /data/mirror/gpseg3 41003 Passive Resynchronizing
  15.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:- gw_sdw1 /data/mirror/gpseg4 41000 Passive Resynchronizing
  16.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:- gw_sdw1 /data/mirror/gpseg5 41001 Passive Resynchronizing
  17.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:- gw_sdw1 /data/mirror/gpseg6 41002 Passive Resynchronizing
  18.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:- gw_sdw1 /data/mirror/gpseg7 41003 Passive Resynchronizing
  19.  
    20190326:01:09:51:031359 gpstate:gw_mdw1:gpadmin-[INFO]:--------------------------------------------------------------

查看节点状态

发现所有的mirror已经启动

  1.  
    postgres=# SELECT * from gp_segment_configuration ;
  2.  
    dbid | content | role | preferred_role | mode | status | port | hostname | address | replication_port | san_mounts
  3.  
    ------+---------+------+----------------+------+--------+-------+----------+---------+------------------+------------
  4.  
    1 | -1 | p | p | s | u | 5432 | gw_mdw1 | gw_mdw1 | |
  5.  
    2 | 0 | p | p | s | u | 40000 | gw_sdw1 | gw_sdw1 | 43000 |
  6.  
    10 | 0 | m | m | s | u | 41000 | gw_sdw2 | gw_sdw2 | 42000 |
  7.  
    3 | 1 | p | p | s | u | 40001 | gw_sdw1 | gw_sdw1 | 43001 |
  8.  
    11 | 1 | m | m | s | u | 41001 | gw_sdw2 | gw_sdw2 | 42001 |
  9.  
    4 | 2 | p | p | s | u | 40002 | gw_sdw1 | gw_sdw1 | 43002 |
  10.  
    12 | 2 | m | m | s | u | 41002 | gw_sdw2 | gw_sdw2 | 42002 |
  11.  
    5 | 3 | p | p | s | u | 40003 | gw_sdw1 | gw_sdw1 | 43003 |
  12.  
    13 | 3 | m | m | s | u | 41003 | gw_sdw2 | gw_sdw2 | 42003 |
  13.  
    6 | 4 | p | p | s | u | 40000 | gw_sdw2 | gw_sdw2 | 43000 |
  14.  
    14 | 4 | m | m | s | u | 41000 | gw_sdw1 | gw_sdw1 | 42000 |
  15.  
    7 | 5 | p | p | s | u | 40001 | gw_sdw2 | gw_sdw2 | 43001 |
  16.  
    15 | 5 | m | m | s | u | 41001 | gw_sdw1 | gw_sdw1 | 42001 |
  17.  
    8 | 6 | p | p | s | u | 40002 | gw_sdw2 | gw_sdw2 | 43002 |
  18.  
    16 | 6 | m | m | s | u | 41002 | gw_sdw1 | gw_sdw1 | 42002 |
  19.  
    9 | 7 | p | p | s | u | 40003 | gw_sdw2 | gw_sdw2 | 43003 |
  20.  
    17 | 7 | m | m | s | u | 41003 | gw_sdw1 | gw_sdw1 | 42003 |
  21.  
    (17 rows)

Greenplum 添加mirror步骤的更多相关文章

  1. Deepgreen/Greenplum 删除节点步骤

    Deepgreen/Greenplum删除节点步骤 Greenplum和Deepgreen官方都没有给出删除节点的方法和建议,但实际上,我们可以对节点进行删除.由于不确定性,删除节点极有可能导致其他的 ...

  2. pytest文档38-allure.setp添加测试用例步骤

    前言 一般流程性的测试用例,写成自动化用例时,步骤较多写起来会比较长.在测试用例里面添加详细的步骤有助于更好的阅读,也方便报错后快速的定位到问题. 举个常见的测试场景用例:从登陆开始,到浏览商品添加购 ...

  3. 01 Mybatis框架添加英雄步骤

    客户端发出请求的几种方式 通过浏览器的地址栏中发出请求 通过html页面中的超链接发出请求 通过html页面中的form表单发出请求 通过前端框架发出请求 工程中使用数据库需要做的几件事: 在pom. ...

  4. oneThink后台添加插件步骤详解

    内容管理框架:oneThink 版本:V1.1.141212 (注:v1.1也有很多版本,一不小心就下到V1.1.140202 去了,还有其他版本,建议去代码托管平台下载最新版本) 我也不偷懒,把每一 ...

  5. 【转】Hadoop集群添加磁盘步骤

    转自:http://blog.csdn.net/huyuxiang999/article/details/17691405 一.实验环境 : 1.硬件:3台DELL服务器,CPU:2.27GHz*16 ...

  6. ASP.NET Core 添加区域步骤（详细）

    1 前言 早就想总结一下,但是没时间,这次有时间了,就详细的把步骤写出来. 2 步骤 2.1 添加区域 右键项目->添加->区域,如图1: 区域名称:Ceshi 添加完之后解决方案管理器会 ...

  7. linux操作系统不重启添加raid0步骤

    1.限制:本步骤仅适用于LSI芯片的raid卡,可以通过以下蓝色指令判断是否LSI芯片 [root@HKC-Lab-CDN ~]# lspci | grep -i lsi 03:00.0 RAID b ...

  8. 阿里云移动推送 ios项目添加SDK步骤

    添加阿里云Pods仓库和各产品SDK Pod依赖,配置步骤如下: 1. CocoaPods集成添加阿里云Pods仓库,Podfile添加: source 'https://github.com/ali ...

  9. 阿里云移动推送 安卓项目添加SDK步骤

    Gradle的Emas服务插件会加载您下载的 aliyun-emas-services.json 文件.请修改工程的build.gradle使用该插件.配置步骤如下: 1. 修改项目级目录下build ...

随机推荐

  1. Python脚本-自动下载安装

    #coding=utf-8 import os import sys if os.getuid() == 0: pass else: print 'no' sys.exit(1) version = ...

  2. 【HC89S003F4开发板】 10汇编指令

    HC89S003F4开发板汇编指令 一.数据传递类指令 MOV.MOVC.MOVX 1.MOV,用于片内数据存储器中的数据传递指令中. 2.MOVC是与ROM之间的数据传送,而MOVX是与外部RAM数 ...

  3. Go 协程

    Go 协程 协程与传统的系统级线程和进程相比,协程的优势在于其"轻量级",可以轻松创建上百万个协程而不会导致系统资源衰竭,所以协程也叫做轻量级线程. 在Go中goroutine就是 ...

  4. prometheus grafana graylog 钉钉告警 短信告警 电话告警系统 PrometheusAlert

    PrometheusAlert 简介 PrometheusAlert是开源的运维告警中心消息转发系统,支持主流的监控系统Prometheus,日志系统Graylog和数据可视化系统Grafana发出的 ...

  5. canal+kafka订阅Mysql binlog将数据异构到elasticsearch(或其他存储方式)

    canal本质就是"冒充"从库,通过订阅mysql bin-log来获取数据库的更改信息. mysql配置(my.cnf) mysql需要配置my.cnf开启bin-log日志并且 ...

  6. Codeforces 1207 G. Indie Album

    Codeforces 1207 G. Indie Album 解题思路 离线下来用SAM或者AC自动机就是一个单点加子树求和,套个树状数组就好了,因为这个题广义SAM不能存在 \(len[u] = l ...

  7. go 学习笔记（1）go command

    常用命令 go command [arguments] 1) go build  跨平台编译: env GOOS=linux GOARCH=amd64 go build 2) go install : ...

  8. logback 生成 catalina.base_IS_UNDEFINED 问题处理 &如何在eclipse/idea中添加VM参数

    1>在Eclipse中里设置  windows->preferences->Java->Installed JRES->edit->Default VM Argum ...

  9. Java Web 深入分析（11） JVM(1)

    前言 Java启动后作为一个进程运行在操作系统中,该进程要分配的内存有以下几个: 1.Java堆: 存储java内存区域,堆大小是在jvm启动时就像操作系统申请完成,其中 -Xmx和-Xms 分别表示 ...

  10. 5_PHP数组_3_数组处理函数及其应用_7_数组排列函数

    以下为学习孔祥盛主编的<PHP编程基础与实例教程>(第二版)所做的笔记. 数组排列函数 1. sort() 函数 程序: <?php $array = array("img ...