前言：初次接触GBase 8c，看到官网提示和Oracle兼容，希望做一次去O到GBase 8c的验证测试，测试前先部署了一套主备集群，也希望对此熟悉的朋友能多多指正。

一、安装需求

主备环境各服务器应具有相同体系架构

• 64bit 和 32bit 不能同一集群
• ARM 和 x86 两类系统不能同一集群

1.1 硬件环境要求

主备环境GBase 8c 各服务器应满足以下最低硬件需求，生产环境应根据业务需求适时调整硬件配置。

1.2 操作系统要求

1.2.1 软件环境要求

1.2.2 软件依赖要求

二、集群规划

2.1 主机名称规划

2.2 主机地址规划

2.3 端口号规划

2.4 用户及组规划

2.5 软件目录规划

三、环境准备

3.1 安装软件依赖包

# root用户执行【所有节点】
[root@node-dbxxx ~]# yum install -y bison flex libaio-devel lsb_release patch ncurses-devel bzip2
-- 检查是否已安装
[root@node-dbxxx ~]# rpm -qa --queryformat "%{NAME}-%{VERSION}-%{RELEASE} (%{ARCH})\n" | grep -E "|bison|flex|libaio-devel|lsb_release|patch|ncurses-devel|bzip2"

3.2 关闭防火墙

# root用户执行【所有节点】
-- 停止 firewalld 服务
[root@node-dbxxx ~]# systemctl stop firewalld.service
-- 禁用 firewalld 服务
[root@node-dbxxx ~]# systemctl disable firewalld.service
-- 查看 firewalld 服务状态
[root@node-dbxxx ~]# systemctl status firewalld
-- 显示如下表示已关闭禁用防火墙
● firewalld.service - firewalld - dynamic firewall daemon
   Loaded: loaded (/usr/lib/systemd/system/firewalld.service; disabled; vendor preset: enabled)
   Active: inactive (dead)
     Docs: man:firewalld(1)

3.3 关闭SELinux

# root用户执行【所有节点】
-- 临时关闭SELinux
[root@node-dbxxx ~]# setenforce 0
-- 永久关闭SELinux
[root@node-dbxxx ~]# sed -i "s/SELINUX=enforcing/SELINUX=disabled/" /etc/selinux/config
-- 检查SELinux状态
[root@node-dbxxx ~]# getenforce
-- 状态为Disabled表明SELinux已关闭
Disabled

3.4 设置时区和时间

GBase 8c 主备系统需主备节点完成时间同步，保证主备数据的一致性，通常采用NTP服务保证各节点间时间同步。

# root用户执行【所有节点】
# 使用ntp设置时钟同步
-- 安装ntp服务
[root@node-dbxxx ~]# yum install -y ntp
-- 开机启动ntp服务
[root@node-dbxxx ~]# systemctl enable ntpd
-- 启动ntp服务
[root@node-dbxxx ~]# systemctl start ntpd
-- 设置时区Asia/Shanghai
[root@node-dbxxx ~]# timedatectl set-timezone Asia/Shanghai
-- 检查时区
[root@node-dbxxx ~]# timedatectl |grep -i zone 
-- 启用ntp服务
[root@node-dbxxx ~]# timedatectl set-ntp yes
-- 编辑定时任务列表
[root@node-dbxxx ~]# crontab -e
-- 使用vi/vim对定时任务列表进行编辑，添加如下内容变保存退出
0 12 * * * ntpdate cn.pool.ntp.org
-- 查看时间及时区
[root@node-dbxxx ~]# timedatectl status

3.5 修改硬件时钟

硬件时钟是指主机板上的时钟设备，也就是通常可在BIOS画面设定的时钟，服务器启动时会根据硬件时钟设置系统时钟。

# root用户执行【所有节点】
-- 将当前系统时间写入硬件时钟，/etc/adjtime随之更新
[root@node-dbxxx ~]# hwclock --systohc
-- 查看硬件时钟
[root@node-dbxxx ~]# hwclock --show

3.6 配置SSH

# root用户执行【所有节点】
-- 禁用 SSH 登录时的横幅（Banner）
[root@node-dbxxx ~]#sed -i '/Banner/s/^/#/' /etc/ssh/sshd_config
-- 目的是禁用允许以 root 用户登录的配置选项
[root@node-dbxxx ~]# sed -i '/PermitRootLogin/s/^/#/' /etc/ssh/sshd_config
[root@node-dbxxx ~]# echo -e "\n" >> /etc/ssh/sshd_config
[root@node-dbxxx ~]# echo "Banner none " >> /etc/ssh/sshd_config
# 修改Banner配置，去掉连接到系统时，系统提示的欢迎信息。欢迎信息会干扰安装时远程操作的返回结果，影响安装正常执行
[root@node-dbxxx ~]# echo "PermitRootLogin yes" >> /etc/ssh/sshd_config
[root@node-dbxxx ~]# cat /etc/ssh/sshd_config |grep -v ^#|grep -E 'PermitRootLogin|Banner'
-- 重启生效
[root@node-dbxxx ~]# systemctl restart sshd.service
-- 查看SSH状态
[root@node-dbxxx ~]# systemctl status sshd.service

3.7 添加hosts解析

# root用户执行【所有节点】
[root@node-dbxxx ~]# cat >> /etc/hosts<<EOF
10.110.3.153  node-db2  
10.110.3.154  node-db3 
EOF

3.8 调整系统内核参数

如下系统内核参数预安装时脚本未进行自动设置，需通过手工调整。

# root用户执行【所有节点】
[root@node-dbxxx ~]# tee -a /etc/sysctl.conf << EOF
kernel.sem = 40960 2048000 40960 20480
EOF
-- 然后执行sysctl -p使修改生效
[root@node-dbxxx ~]# sysctl -p
-- 查看修改后的kernel.sem值
[root@node-dbxxx ~]# sysctl kernel.sem

3.9 创建用户配置sudoer

# root用户执行【所有节点】
[root@node-dbxxx ~]# groupadd gbase -g 1005
[root@node-dbxxx ~]# useradd gbase -u 1005 -g gbase
[root@node-dbxxx ~]# echo "passwd" | passwd --stdin gbase

-- 配置sudoer
-- 1) 使用visudo命令打开/etc/sudoers文件
[root@node-dbxxx ~]# visudo
-- 2) 在文件中找到以下内容（或类似内容）
## Allow root to run any commands anywhere
-- 3) 将光标移到该行末尾，然后按下 o 键（小写字母 o，不是零）
-- 4) 在新行中输入如下新的内容
gbase ALL=(ALL) NOPASSWD:ALL
-- 5) 按下 Esc 键退出插入模式,然后按键“:wq!”或"zz"保存并退出

3.10 创建目录

# gbase用户执行【主节点】
[gbase@node-db2 ~]$ mkdir -p /home/gbase/gbase_package

# root用户执行【所有节点】
[root@node-dbxxx ~]# mkdir -p /opt/gbase
[root@node-dbxxx ~]# chown gbase:gbase /opt/gbase

3.11 配置互信

# gbase用户执行【主节点】
[gbase@node-db2 ~]$ tee -a /home/gbase/sshtrust.sh << EOF
#!/bin/bash

# 配置用户名和密码
username="gbase"
password="gbase"

# 配置服务器IP列表

servers=(
"10.110.3.153"
"10.110.3.154"
)

# 创建脚本日志文件
log_file="$(dirname "$0")/script.log"
echo "" > "$log_file"

# 生成本地SSH密钥对
if [ ! -f "/home/$username/.ssh/id_rsa" ]; then
  ssh-keygen -q -t rsa -N "" -f "/home/$username/.ssh/id_rsa"
fi

# 拷贝本地SSH公钥到目标服务器
for server in "${servers[@]}"
do
  # 进行互信
  sshpass -p "$password" ssh-copy-id -o StrictHostKeyChecking=no "$username@$server"

  # 检查互信是否成功，并记录日志
  if [ "$?" -eq "0" ]
  then
    echo "$(date +'%Y-%m-%d %H:%M:%S') - 互信成功 - $server" >> "$log_file"

    # 拷贝本地.ssh目录内容到目标服务器
    sshpass -p "$password" scp -r /home/"$username"/.ssh/* "$username@$server":/home/"$username"/.ssh
  else
    echo "$(date +'%Y-%m-%d %H:%M:%S') - 互信失败 - $server" >> "$log_file"
  fi
done
EOF
-- 2) 授权/home/gbase/sshtrust.sh
[gbase@node-db2 ~]$ chmod +x /root/sshtrust.sh
-- 3) 执行/root/sshtrust.sh
[gbase@node-db2 ~]$ sh sshtrust.sh 
/bin/ssh-copy-id: INFO: Source of key(s) to be installed: "/home/gbase/.ssh/id_rsa.pub"
/bin/ssh-copy-id: INFO: attempting to log in with the new key(s), to filter out any that are already installed
/bin/ssh-copy-id: INFO: 1 key(s) remain to be installed -- if you are prompted now it is to install the new keys

Number of key(s) added: 1

Now try logging into the machine, with:   "ssh -o 'StrictHostKeyChecking=no' 'gbase@10.110.3.153'"
and check to make sure that only the key(s) you wanted were added.

/bin/ssh-copy-id: INFO: Source of key(s) to be installed: "/home/gbase/.ssh/id_rsa.pub"
/bin/ssh-copy-id: INFO: attempting to log in with the new key(s), to filter out any that are already installed
/bin/ssh-copy-id: INFO: 1 key(s) remain to be installed -- if you are prompted now it is to install the new keys

Number of key(s) added: 1

Now try logging into the machine, with:   "ssh -o 'StrictHostKeyChecking=no' 'gbase@10.110.3.154'"
and check to make sure that only the key(s) you wanted were added.

3.12 下载软件安装包

3.12.1 下载安装包

使用注册账号登录GBase官网https://www.gbase.cn/download/gbase-8c?category=INSTALL_PACKAGE下载页面，下载与操作系统匹配的[GBase8cV5_S3.0.0_单机主备版_x86_64.tar.zip]软件安装包并将下载的软件包使用gbase用户上传至服务器/home/gbase/gbase_package目录下。

3.12.2 校验安装包

# gbase用户执行【主节点】
[gbase@node-db2 ~]$ cd /home/gbase/gbase_package
[gbase@node-db2 gbase_package]$ unzip GBase8cV5_S3.0.0_单机主备版_x86_64.tar.zip 
[gbase@node-db2 gbase_package]$ ll
total 525488
-rw-rw-r-- 1 gbase gbase 269042643 May  6 10:26 GBase8cV5_S3.0.0B69_centos7.8_x86_64.tar.gz
-rw-r--r-- 1 gbase gbase 269043107 Jul 24 11:33 GBase8cV5_S3.0.0_单机主备版_x86_64.tar.zip
-rw-rw-r-- 1 gbase gbase       133 May  6 10:24 md5.txt

-- 解压下载的GBase8cV5_S3.0.0_单机主备版_x86_64.tar.zip安装包后，里面含有一个md5.txt的文件，内含软件安装包GBase8cV5_S3.0.0B69_centos7.8_x86_64.tar.gz的md5值，使用如下命令校验并对比下载的安装包是否损坏
[gbase@node-db2 gbase_package]$ md5sum GBase8cV5_S3.0.0B69_centos7.8_x86_64.tar.gz
506fe679e7a6d851456d27fe39730f95  GBase8cV5_S3.0.0B69_centos7.8_x86_64.tar.gz

[gbase@node-db2 gbase_package]$ cat md5.txt 
506fe679e7a6d851456d27fe39730f95  GBase8cV5_S3.0.0B69_centos7.8_x86_64.tar.gz
a8badbbd317ed021518d195ee00f8c40b0bc3724        git commit id 

3.12.3 解压安装包

# gbase用户执行【主节点】
[gbase@node-db2 gbase_package]$ tar -zxvf GBase8cV5_S3.0.0B69_centos7.8_x86_64.tar.gz
GBase8cV5_S3.0.0B69_CentOS_x86_64_om.sha256
GBase8cV5_S3.0.0B69_CentOS_x86_64_om.tar.gz
GBase8cV5_S3.0.0B69_CentOS_x86_64_pgpool.tar.gz
GBase8cV5_S3.0.0B69_CentOS_x86_64.sha256
GBase8cV5_S3.0.0B69_CentOS_x86_64.tar.bz2

-- 继续解压GBase8cV5_S3.0.0B69_CentOS_x86_64_om.tar.gz
[gbase@node-db2 gbase_package]$ tar -zxvf GBase8cV5_S3.0.0B69_CentOS_x86_64_om.tar.gz

3.13 编辑yml文件

编辑集群配置文件gbase.yml。

# gbase用户执行【主节点】
-- 部署主备式集群要制定cluster_type参数为 single-inst
[gbase@node-db2 gbase_package]$ cp -p gbase.yml /home/gbase/
[gbase@node-db2 gbase_package]$ cd ~
[gbase@node-db2 ~]$ vim gbase.yml
-- 修改后的gbase.yml信息如下
gha_server:
  - gha_server1:
      host: 10.110.3.153
      port: 20001
dcs:
  - host: 10.110.3.153
    port: 2379
  - host: 10.110.3.154
    port: 2379

datanode:
  - dn1:
      - dn1_1:
          host: 10.110.3.153
          agent_host: 10.110.3.153
          role: primary
          port: 15432
          agent_port: 8005
          work_dir: /opt/gbase/data/dn/dn_1
      - dn1_2:
          host: 10.110.3.154
          agent_host: 10.110.3.154
          role: standby
          port: 15433
          agent_port: 8006
          work_dir: /opt/gbase/data/dn/dn_2

env:
  # cluster_type allowed values: multiple-nodes, single-inst, default is multiple-nodes
  cluster_type: single-inst
  pkg_path: /home/gbase/gbase_package
  prefix: /home/gbase/gbase_db
  version: V5_S3.0.0B69
  user: gbase
  port: 22
  third_ssh: false
# constant:
#   virtual_ip: 100.0.1.254/24

四、执行安装

4.1 安装数据库

# gbase用户执行【主节点】
[gbase@node-db2 ~]$ cd /home/gbase/gbase_package/script
[gbase@node-db2 script]$ ./gha_ctl install -p /home/gbase/ -c gbase              
{
    "ret":0,
    "msg":"Success"
}

-- 返回上述内容即表示安装成功

-- 执行过程中可以进入到集群各节点/home/gbase/gbase_db/log/om目录下查看安装日志
[root@node-db2 om]# pwd
/home/gbase/gbase_db/log/om
[root@node-db2 om]# ls -l
total 24
-rw------- 1 gbase gbase  1739 Jul 25 09:35 gs_install-2023-07-25_093427.log
-rw------- 1 gbase gbase 19707 Jul 25 09:36 gs_local-2023-07-25_093429.log

4.2 安装验证

4.2.1 检查数据库集群状态

# gbase用户执行【任意节点】
[gbase@node-dbxxx ~]$ gha_ctl monitor all -l http://10.110.3.153:2379 -H
+----+-------------+--------------+-------+---------+--------+
| No |     name    |     host     |  port |  state  | leader |
+----+-------------+--------------+-------+---------+--------+
| 0  | gha_server1 | 10.110.3.153 | 20001 | running |  True  |
+----+-------------+--------------+-------+---------+--------+
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| No | group |  name |     host     |  port |         work_dir        |  state  |   role  |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| 0  |  dn1  | dn1_1 | 10.110.3.153 | 15432 | /opt/gbase/data/dn/dn_1 | running | primary |
| 1  |  dn1  | dn1_2 | 10.110.3.154 | 15433 | /opt/gbase/data/dn/dn_2 | running | standby |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
+----+--------------------------+--------+---------+----------+
| No |           url            |  name  |  state  | isLeader |
+----+--------------------------+--------+---------+----------+
| 0  | http://10.110.3.153:2379 | node_0 | healthy |  False   |
| 1  | http://10.110.3.154:2379 | node_1 | healthy |   True   |
+----+--------------------------+--------+---------+----------+

4.2.2 启停集群测试

4.2.2.1 停止集群

# gbase用户执行【任意节点】
[gbase@node-dbxxx ~]$ gha_ctl stop all -l http://10.110.3.153:2379
{
    "ret":0,
    "msg":"Success"
}

-- 查看集群状态
[gbase@node-dbxxx ~]$ gha_ctl monitor all -l http://10.110.3.153:2379 -H
+----+-------------+--------------+-------+---------+--------+
| No |     name    |     host     |  port |  state  | leader |
+----+-------------+--------------+-------+---------+--------+
| 0  | gha_server1 | 10.110.3.153 | 20001 | stopped | False  |
+----+-------------+--------------+-------+---------+--------+
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| No | group |  name |     host     |  port |         work_dir        |  state  |   role  |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| 0  |  dn1  | dn1_1 | 10.110.3.153 | 15432 | /opt/gbase/data/dn/dn_1 | stopped | primary |
| 1  |  dn1  | dn1_2 | 10.110.3.154 | 15433 | /opt/gbase/data/dn/dn_2 | stopped | standby |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
+----+--------------------------+--------+---------+----------+
| No |           url            |  name  |  state  | isLeader |
+----+--------------------------+--------+---------+----------+
| 0  | http://10.110.3.153:2379 | node_0 | healthy |  False   |
| 1  | http://10.110.3.154:2379 | node_1 | healthy |   True   |
+----+--------------------------+--------+---------+----------+

4.2.2.2 启动集群

# gbase用户执行【任意节点】
[gbase@node-db3 ~]$ gha_ctl start all -l http://10.110.3.153:2379
{
    "ret":0,
    "msg":"Success"
}

-- 查看集群状态
[gbase@node-db3 ~]$ gha_ctl monitor all -l http://10.110.3.153:2379 -H
+----+-------------+--------------+-------+---------+--------+
| No |     name    |     host     |  port |  state  | leader |
+----+-------------+--------------+-------+---------+--------+
| 0  | gha_server1 | 10.110.3.153 | 20001 | running |  True  |
+----+-------------+--------------+-------+---------+--------+
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| No | group |  name |     host     |  port |         work_dir        |  state  |   role  |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| 0  |  dn1  | dn1_1 | 10.110.3.153 | 15432 | /opt/gbase/data/dn/dn_1 | running | primary |
| 1  |  dn1  | dn1_2 | 10.110.3.154 | 15433 | /opt/gbase/data/dn/dn_2 | running | standby |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
+----+--------------------------+--------+---------+----------+
| No |           url            |  name  |  state  | isLeader |
+----+--------------------------+--------+---------+----------+
| 0  | http://10.110.3.153:2379 | node_0 | healthy |  False   |
| 1  | http://10.110.3.154:2379 | node_1 | healthy |   True   |
+----+--------------------------+--------+---------+----------+

# 集群在启动过程中，可以看到集群的状态state是从stopped--> unstable --> running

4.3.3 连接数据库测试

# gbase用户执行【任意节点】
[gbase@node-db2 ~]$ gsql -d postgres -p 15432
gsql ((single_node GBase8cV5 3.0.0B69 build a8badbbd) compiled at 2023-03-31 15:23:20 commit 0 last mr 1110 )
Non-SSL connection (SSL connection is recommended when requiring high-security)
Type "help" for help.

postgres=# \l
                         List of databases
   Name    | Owner | Encoding | Collate | Ctype | Access privileges 
-----------+-------+----------+---------+-------+-------------------
 postgres  | gbase | UTF8     | C       | C     | 
 template0 | gbase | UTF8     | C       | C     | =c/gbase         +
           |       |          |         |       | gbase=CTc/gbase
 template1 | gbase | UTF8     | C       | C     | =c/gbase         +
           |       |          |         |       | gbase=CTc/gbase
(3 rows)

postgres=# create database gbasedb;
CREATE DATABASE
postgres=# \q

[gbase@node-db2 ~]$ gsql -d postgres -p 15433
gsql ((single_node GBase8cV5 3.0.0B69 build a8badbbd) compiled at 2023-03-31 15:23:20 commit 0 last mr 1110 )
Non-SSL connection (SSL connection is recommended when requiring high-security)
Type "help" for help.

postgres=# \l
                         List of databases
   Name    | Owner | Encoding | Collate | Ctype | Access privileges 
-----------+-------+----------+---------+-------+-------------------
 gbasedb   | gbase | UTF8     | C       | C     | 
 postgres  | gbase | UTF8     | C       | C     | 
 template0 | gbase | UTF8     | C       | C     | =c/gbase         +
           |       |          |         |       | gbase=CTc/gbase
 template1 | gbase | UTF8     | C       | C     | =c/gbase         +
           |       |          |         |       | gbase=CTc/gbase
(4 rows)

postgres=# create database gbase8c;
CREATE DATABASE

五、附录

5.1 “msg”:“Resource:gbase already in use”

问题描述：

在安装GBase主备集群时，第一次未安装成功，当再次安装时，总是报如下信息，无法继续安装：

[gbase@node-db2 script]$ ./gha_ctl install -p /home/gbase/ -c gbase
{
    "ret":80000107,
    "msg":"Resource:gbase already in use"
}

即使删除主备节点/home/gbase/gbase_db及/tmp/gha_ctl目录也还是无法安装

解决办法：

使用gbase用户进入到主节点/home/gbase/gbase_package/scripts目录下执行如下命令进行卸载DCS工具

[gbase@node-db2 ~]$ cd /home/gbase/gbase_package/script/
[gbase@node-db2 script]$ ./gha_ctl uninstall -c gbase -l http://10.110.3.153:2379
{
    "ret":0,
    "msg":"Success"
}
-- 当出现如上信息时提示卸载成功
-- 实际生产测试中

当DCS工具卸载干净，可以重新进行安装。

5.2 主备集群状态正常时执行集群启动任务竟然会先停集群再启动

在主备集群安装完成后，我尝试执行gha_ctl start all -l http://10.110.3.153:2379 启动集群，结果发现集群状态先是被停止了，然后又将集群重新进行了启动。

[gbase@node-db2 ~]$ gha_ctl start all -l http://10.110.3.153:2379
{
    "ret":0,
    "msg":"Success"
}
-- 此时新开一个会话监控集群状态，会发现集群状态state先stopped
[gbase@node-db2 ~]$ gha_ctl monitor all -l http://10.110.3.153:2379 -H
+----+-------------+--------------+-------+---------+--------+
| No |     name    |     host     |  port |  state  | leader |
+----+-------------+--------------+-------+---------+--------+
| 0  | gha_server1 | 10.110.3.153 | 20001 | stopped | False  |
+----+-------------+--------------+-------+---------+--------+
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| No | group |  name |     host     |  port |         work_dir        |  state  |   role  |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| 0  |  dn1  | dn1_1 | 10.110.3.153 | 15432 | /opt/gbase/data/dn/dn_1 | stopped | primary |
| 1  |  dn1  | dn1_2 | 10.110.3.154 | 15433 | /opt/gbase/data/dn/dn_2 | stopped | standby |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
+----+--------------------------+--------+---------+----------+
| No |           url            |  name  |  state  | isLeader |
+----+--------------------------+--------+---------+----------+
| 0  | http://10.110.3.153:2379 | node_0 | healthy |  False   |
| 1  | http://10.110.3.154:2379 | node_1 | healthy |   True   |
+----+--------------------------+--------+---------+----------+

-- 一段时间后，集群状态显示为running
[gbase@node-db2 ~]$ gha_ctl monitor all -l http://10.110.3.153:2379 -H
+----+-------------+--------------+-------+---------+--------+
| No |     name    |     host     |  port |  state  | leader |
+----+-------------+--------------+-------+---------+--------+
| 0  | gha_server1 | 10.110.3.153 | 20001 | running |  True  |
+----+-------------+--------------+-------+---------+--------+
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| No | group |  name |     host     |  port |         work_dir        |  state  |   role  |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
| 0  |  dn1  | dn1_1 | 10.110.3.153 | 15432 | /opt/gbase/data/dn/dn_1 | running | primary |
| 1  |  dn1  | dn1_2 | 10.110.3.154 | 15433 | /opt/gbase/data/dn/dn_2 | running | standby |
+----+-------+-------+--------------+-------+-------------------------+---------+---------+
+----+--------------------------+--------+---------+----------+
| No |           url            |  name  |  state  | isLeader |
+----+--------------------------+--------+---------+----------+
| 0  | http://10.110.3.153:2379 | node_0 | healthy |  False   |
| 1  | http://10.110.3.154:2379 | node_1 | healthy |   True   |
+----+--------------------------+--------+---------+----------+

我觉得这点不太合理，在集群状态正常情况下再启动集群应该提示集群已启动，而不是重启，此问题已反馈GBase公司相应人员。