背景信息

其中替换服务器根据使用场景不同，可分为如下两种：

• 服务器替换前后IP一致：将新服务器的名称、IP与故障服务器的名称、IP修改一致，然后做替换。适用于用户保证替换服务器的名称、IP和被替换服务器的名称、IP一致的场景，使用此方案时要求的前置条件少，能快速实现替换。
• 服务器替换前后IP不一致：无需修改新服务器的名称、IP，直接做替换。适用于替换服务器的名称、IP和被替换服务器的名称、IP不一致的场景，也可用于将一个高性能的机器替换一个性能较差但其上实例状态正常的场景，使用此方案时要求的前置条件较多。

“不换IP”和“换IP”是从集群配置角度的差异，“不换IP”指不修改集群配置文件的IP和hostname信息，“换IP”是指需要修改集群配置文件的IP和hostname信息。

前提条件

• 分布式部署模式下替换故障实例
  • 故障服务器不能存在CN实例。
  • 集群状态可查，主CM Server状态正常，集群状态不能为Unavailable。
  • 替换前不能锁定集群。
  • 实例替换必须在一个没有完全损坏的服务器上执行。
  • 在替换实例时，各服务器必须互信正常。
  • 替换实例操作会修复故障服务器下的全部故障实例。
  • 替换实例期间集群不能执行DDL和DML操作。
  • 集群状态必须异常。
• 分布式部署模式下替换服务器（不换IP）
  • 故障服务器不能存在CN实例。
  • 集群状态可查，主CM Server状态正常，集群状态不能为Unavailable。
  • 替换前不能锁定集群。
  • 节点替换必须在一个没有完全损坏的服务器上执行。
  • 在节点替换前，基于当前集群配置文件执行前置成功。
  • 新服务器的IP、主机名等信息要和损坏的服务器一致。
  • 需保证新服务器和集群其他服务器时间一致。
  • 替换服务器期间集群不能执行DDL和DML操作。
  • 已经卸载故障节点。
  • 集群状态必须异常。
• 分布式部署模式下替换服务器（换IP）

  替换主机-换IP的方案相当于集群温备方案，即无需修改新机IP、hostname的情况下，将处于运行态的服务器直接替换为新服务器。

  • 故障服务器不能存在CN实例。
  • 集群状态可查，主CM Server状态正常，集群状态不能为Unavailable。
  • 替换前不能锁定集群。
  • 节点替换必须在一个没有完全损坏的服务器上执行。
  • 新服务器已经准备好。
  • 需保证新服务器和集群其他服务器时间一致。
  • 基于新服务器IP的集群配置文件已准备好，与老的集群配置文件相比，仅修改了IP和hostname。
  • 在节点替换（换IP）前，基于新集群配置文件执行前置成功。
  • 在执行过程中不能执行DDL和DML语句。
  • 已经卸载故障节点。
  • 集群状态必须异常。
• 主备部署模式下替换故障实例
  • 集群状态可查，且主CM Server状态正常。
  • 替换前，故障服务器上的DN所在的group不能被锁定。
  • 实例替换必须在一个没有完全损坏的服务器上执行。
  • 在替换实例时，各服务器必须互信正常。
  • 替换实例操作会修复故障节点下的全部故障实例。
  • 替换实例期间故障服务器上的DN所在的group不能执行DDL和DML操作。
  • 集群状态必须异常。
• 主备部署模式下替换节点（不换IP）
  • 集群状态可查，且主CM Server状态正常。
  • 替换前，故障服务器上的DN所在的group不能被锁定。
  • 实例替换必须在一个没有完全损坏的主机上执行。
  • 新服务器的IP、服务器名等信息要和损坏的服务器一致。
  • 需保证新服务器和集群其他服务器时间一致。
  • 在节点替换前，基于当前集群配置文件执行前置成功。
  • 替换节点期间故障服务器上的DN所在的group不能执行DDL和DML操作。
  • 已经卸载故障节点。
  • 集群状态必须异常。
• 主备部署模式下替换节点（换IP）
  • 集群状态可查，且主CM Server状态正常。
  • 节点替换必须在一个没有完全损坏的服务器上执行。
  • 新服务器已经准备好。
  • 需保证新服务器和集群其他服务器时间一致。
  • 集群状态可查，且主CM Server状态正常。
  • 基于新服务器IP的集群配置文件已准备好，与老的集群配置文件相比，仅修改了IP和hostname。
  • 在节点替换（换IP）前，基于新集群配置文件执行前置成功。
  • 替换节点期间故障服务器上的DN所在的group不能执行DDL和DML操作。
  • 已经卸载故障节点。
  • 集群状态必须异常。

语法

• 新增服务器

  gs_replace -t install -h HOSTNAME [-l LOGFILE]

  配置实例、服务器

• gs_replace -t config -h HOSTNAME [-l LOGFILE]

  启动新增服务器

• gs_replace -t start -h HOSTNAME [-l LOGFILE]

  替换服务器（换IP）

• gs_replace -t warm-standby  -X XMLFILE [-l LOGFILE]

  显示帮助信息

• gs_replace { -? | --help }

  显示版本号信息

• gs_replace { -V | --version }

  参数说明
  

gs_replace参数可以分为如下几类：

• 通用参数：
  • -t

    om命令的类型。

    取值范围：warm-standby、install、config和start.

    

    进行过主备切换后，如果当前集群状态中balanced字段为false，再继续执行warm-standby，集群状态中balanced字段会显示为true，但实际集群DN主备状态并未改变。

  • -l

    指定日志文件及存放路径。

    默认值：$GAUSSLOG/om/gs_replace.log

• 安装新增实例、主机参数：

  -h

  指定新增服务器或新增实例所在服务器的名称。可以指定多个服务器名称，以逗号分隔。

  取值范围：服务器名称。

• 配置新增实例、主机参数：

  -h

  指定替换服务器或替换实例所在服务器的名称。可以指定多个服务器名称，以逗号分隔。

  取值范围：服务器名称。

• 启动新增实例、服务器参数：

  -h

  指定需要启动服务器或实例所在服务器的名称。可以指定多个服务器名称，以逗号分隔。

  取值范围：服务器名称。

• 替换节点（换IP）参数：
  • -X

    集群配置文件路径。

    取值范围：clusterconfig.xml的路径。

  • -l

    指定日志文件及存放路径。

    在内部会自动给日志名添加一个时间戳。

    • 当既不明确指定-l，又不在XML文件中配置gaussdbLogPath时，默认值为“$GAUSSLOG/用户名/om/gs_replace.log”。
    • 当不明确指定-l，但在XML文件中配置了gaussdbLogPath时，默认值为“gaussdbLogPath的值、用户名和om/gs_replace.log”的组合。
• 其他参数：
  • -?，--help

    显示帮助信息。

  • -V，--version

    显示版本号信息。

示例

• 替换故障实例。
  1. 配置替换的实例。

     omm@plat1:/opt/software/gaussdb/script> gs_replace -t config -h plat1
     Check cluster status for replace.
     Check the status of ETCD cluster.
     Successfully check the status of ETCD cluster.
     Successfully check cluster status.
     Filter out all valid hosts for replacing.
     Distributing configuration to remote host.
     Successfully distributed configuration to remote host.
     Configuring
     Stopping replace instances.
     Successfully stopped replace instances.
     Configuring replacement instances.
     Config replace instances.
     Successfully config replace instances.
     Config standby DN of new instances.
     .......67s
     Config standby DN of new instances successfully.
     Successfully configured replacement instances.
     Configuration succeeded.
     ==============================
     Time statistics:
             Config replacement nodes: 95s
     total: 95s

  2. 启动替换实例所在服务器。

     omm@plat1:/opt/software/gaussdb/script> gs_replace -t start -h plat1
     Starting.
     ==============================
     ..2s
     Start cm agent on new nodes.
     Successfully start cm agent for new nodes.
     Starting the cluster.
     .1s
     Successfully started instance process. Waiting to become Normal.
     ==============================
     Successfully started cluster.
     ==============================
     Time statistics:
             Start replacement nodes: 27s
     total: 27s

• 替换服务器（不换IP）。
  1. 在故障节点执行卸载当前节点命令，卸载故障节点。

     omm@plat4:/opt/software/gaussdb/script> gs_uninstall --delete-data -L
     Check preinstall on every node.
     Successfully checked preinstall on every node.
     Stop cluster.
     Check logfile path.
     Clean crontab.
     Clean crontab successfully.
     Kill process for components.
     Kill process for components successfully.
     Uninstall components
     Uninstall components successfully.
     Modifying user's environmental variable.
     Successfully modified user's environmental variable.
     Clean tmp files and logs.
     Successfully clean cluster's tmp and logs.
     Successful uninstallation

  2. 在正常节点上准备集群环境。

     plat1:/opt/software/gaussdb/script # ./gs_preinstall -U omm -G dbgrp -X /opt/software/gaussdb/clusterconfig.xml
     Parsing the configuration file.
     Successfully parsed the configuration file.
     Installing the tools on the local node.
     Successfully installed the tools on the local node.
     Are you sure you want to create trust for root (yes/no)? yes
     Please enter password for root.
     Password: 
     Creating SSH trust for the root permission user.
     Successfully created SSH trust for the root permission user.
     Pass over configuring LVM
     Distributing package.
     Successfully distributed package.
     Are you sure you want to create the user[omm] and create trust for it (yes/no)? yes
     Please enter password for cluster user.
     Password: 
     Please enter password for cluster user again.
     Password: 
     Creating [omm] user on all nodes.
     Successfully created [omm] user on all nodes.
     Installing the tools in the cluster.
     Successfully installed the tools in the cluster.
     Checking hostname mapping.
     Successfully checked hostname mapping.
     Creating SSH trust for [omm] user.
     Start creating SSH trust for [omm] user.
     Successfully created SSH trust for [omm] user.
     Checking OS version.
     Successfully checked OS version.
     Creating cluster's path.
     Successfully created cluster's path.
     Setting SCTP service.
     Successfully set SCTP service.
     Set and check OS parameter.
     Setting OS parameters.
     Successfully set OS parameters.
     Set and check OS parameter completed.
     Preparing CRON service.
     Successfully prepared CRON service.
     Preparing SSH service.
     Successfully prepared SSH service.
     Setting user environmental variables.
     Successfully set user environmental variables.
     Configuring alarms on the cluster nodes.
     Successfully configured alarms on the cluster nodes.
     Setting the dynamic link library.
     Successfully set the dynamic link library.
     Fixing server package owner.
     Successfully fixed server package owner.
     Create logrotate service.
     Successfully create logrotate service.
     Setting finish flag.
     Successfully set finish flag.
     Clean SSH trust for the root permission user.
     Successfully cleaned SSH trust for the root permission user.
     Preinstallation succeeded.

  3. 在正常节点上安装新增的服务器。

     omm@plat1:/opt/software/gaussdb/script> gs_replace -t install -h plat3
     Distributing configuration to remote host.
     Successfully distributed configuration to remote host.
     Installing.
     Checking installation environment on nodes.
     Successfully checking installation environment on nodes.
     Installing applications on nodes.
     Installation is completed.
     ==============================
     Time statistics:
             Install replacement nodes: 9s
     total: 9s

  4. 配置新增的服务器。

     omm@plat1:/opt/software/gaussdb/script> gs_replace -t config -h plat3
     Check cluster status for replace.
     Check the status of ETCD cluster.
     Successfully check the status of ETCD cluster.
     Successfully check cluster status.
     Filter out all valid hosts for replacing.
     Distributing configuration to remote host.
     Successfully distributed configuration to remote host.
     Configuring
     Stopping replace instances.
     Successfully stopped replace instances.
     Configuring replacement instances.
     Config replace instances.
     Successfully config replace instances.
     Config standby DN of new instances.
     .......67s
     Config standby DN of new instances successfully.
     Successfully configured replacement instances.
     Configuration succeeded.
     ==============================
     Time statistics:
             Config replacement nodes: 95s
     total: 95s

  5. 启动新增的服务器。

     omm@plat1:/opt/software/gaussdb/script> gs_replace -t start -h plat3
     Starting.
     ==============================
     ..2s
     Start cm agent on new nodes.
     Successfully start cm agent for new nodes.
     Starting the cluster.
     ......6s
     Successfully started instance process. Waiting to become Normal.
     ==============================
     Successfully started cluster.
     ==============================
     Time statistics:
             Start replacement nodes: 35s
     total: 35s

• 替换服务器（换IP）。
  1. 在故障节点执行卸载当前节点命令，卸载故障节点。

     omm@plat4:/opt/software/gaussdb/script> gs_uninstall --delete-data -L
     Check preinstall on every node.
     Successfully checked preinstall on every node.
     Stop cluster.
     Check logfile path.
     Clean crontab.
     Clean crontab successfully.
     Kill process for components.
     Kill process for components successfully.
     Uninstall components
     Uninstall components successfully.
     Modifying user's environmental variable.
     Successfully modified user's environmental variable.
     Clean tmp files and logs.
     Successfully clean cluster's tmp and logs.
     Successful uninstallation

  2. 在集群配置文件中将plat4替换为plat5，然后执行如下命令完成前置操作。

     plat1:/opt/software/gaussdb/script # ./gs_preinstall -U omm -G dbgrp -X /opt/software/gaussdb/clusterconfig.xml
     Parsing the configuration file.
     Successfully parsed the configuration file.
     Installing the tools on the local node.
     Successfully installed the tools on the local node.
     Are you sure you want to create trust for root (yes/no)? yes
     Please enter password for root.
     Password: 
     Creating SSH trust for the root permission user.
     Successfully created SSH trust for the root permission user.
     Pass over configuring LVM
     Distributing package.
     Successfully distributed package.
     Are you sure you want to create the user[omm] and create trust for it (yes/no)? yes
     Please enter password for cluster user.
     Password: 
     Please enter password for cluster user again.
     Password: 
     Creating [omm] user on all nodes.
     Successfully created [omm] user on all nodes.
     Installing the tools in the cluster.
     Successfully installed the tools in the cluster.
     Checking hostname mapping.
     Successfully checked hostname mapping.
     Creating SSH trust for [omm] user.
     Start creating SSH trust for [omm] user.
     Successfully created SSH trust for [omm] user.
     Checking OS version.
     Successfully checked OS version.
     Creating cluster's path.
     Successfully created cluster's path.
     Setting SCTP service.
     Successfully set SCTP service.
     Set and check OS parameter.
     Setting OS parameters.
     Successfully set OS parameters.
     Set and check OS parameter completed.
     Preparing CRON service.
     Successfully prepared CRON service.
     Preparing SSH service.
     Successfully prepared SSH service.
     Setting user environmental variables.
     Successfully set user environmental variables.
     Configuring alarms on the cluster nodes.
     Successfully configured alarms on the cluster nodes.
     Setting the dynamic link library.
     Successfully set the dynamic link library.
     Fixing server package owner.
     Successfully fixed server package owner.
     Create logrotate service.
     Successfully create logrotate service.
     Setting finish flag.
     Successfully set finish flag.
     Clean SSH trust for the root permission user.
     Successfully cleaned SSH trust for the root permission user.
     Preinstallation succeeded.

  3. 在正常节点上执行如下命令进行替换服务器操作。

     omm@plat1:/opt/software/gaussdb/script> gs_replace -t warm-standby -X  /opt/software/gaussdb/clusterconfig.xml
     Checking the cluster configuration differences.
     Successfully checked the cluster configuration differences.
     Check cluster status for replace.
     Check the status of ETCD cluster.
     Successfully check the status of ETCD cluster.
     Successfully check cluster status.
     Checking replace condition.
     Successfully checked replace condition.
     Check Preinstall Flag.
     Creating the backup directory.
     Successfully created backup directory.
     Changing replace IP.
     Changing instance configuration files.
     Successfully changed instance configuration files.
     Updating cluster database node info.
     Successfully update cluster database node info.
     Successfully changed replacement IP addresses.
     Install GaussDB 100 on plat11.
     ..............14s
     Successfully install GaussDB 100 on warm-standby nodes.
     Config GaussDB 100 on plat11.
     ..................167s
     Successfully config GaussDB 100 on warm-standby nodes.
     Start GaussDB 100 on plat11.
     ...........54s
     Successfully started warm-standby nodes.
     Warm standby is completed.
     ==============================
     Time statistics:
             Update cluster configuration file.: 381s
     total: 381s

相关命令

gs_preinstall，gs_install，gs_uninstall