lightdb_service.py - lightdb一键启停脚本

lightdb_service.py-lightdb一键启停脚本

在这里对最新版的lightdb_service.py的功能及使用进行介绍

一. 概述

lightdb_service.py 用于对lightdb 服务进行简易的一键启停操作，包括高可用模式，分布式模式和单机模式。

lightdb_service.py 基于python3开发。

二. 用法

usage: lightdb_service.py -c {start,stop,restart,status} [-C | -D] [-P | -S] [-n <node_info>]
                                 [-F <install_info>] [-B <parallel_processes>] [-v] [-q] [-l <directory>]
                                 [-i] [--dry-run] [-f] [-h]
Example:
  start: lightdb_service.py  -c start 
  stop: lightdb_service.py  -c stop 
  restart: lightdb_service.py  -c restart 
  primary only: lightdb_service.py  -c start/stop/restart -P
  standby only: lightdb_service.py  -c start/stop/restart -S
  cn only: lightdb_service.py  -c start/stop/restart -C
  dn only: lightdb_service.py  -c start/stop/restart -D
  dn and primary only: lightdb_service.py  -c start/stop/restart -D -P
  specify node: lightdb_service.py  -c start/stop/restart -n ip:port

Use for start/stop/restart LightDB service

optional arguments:
  -h, --help            show this help message and exit
  -c {start,stop,restart,status}, --command {start,stop,restart,status}
                        start/stop/restart/status lightdb service
  -F <install_info>, --filename <install_info>
                        specifies the path to a json file containing
                        installation information(uninstall/uninstallFile.json)
  -C, --cn_only         only start/stop/restart coordinator node
  -D, --dn_only         only start/stop/restart data node
  -P, --primary_only    only start/stop/restart primary node for primary
                        restart
  -S, --standby_only    only start/stop/restart standby node for standby
                        restart
  -n <node_info>, --node <node_info>
                        only start/stop/restart specified node, node_info
                        formart: ip:port
  --dry-run             show what would happen for action, but don't execute
                        it
  -f, --force           force stop of cluster even if some nodes are in
                        incorrect state, skip incorrect nodes
  -B <parallel_processes>, --parallel <parallel_processes>
                        number of segment hosts to run in parallel. Default is
                        1
  -v, --verbose         debug print
  -q, --quiet           suppress status messages for stdout logging
  -l <directory>, --log_dir <directory>
                        Logfile directory, default is /tmp/ltAdminLogs
  -i, --ignore-keepalived
                        ignore keepalived, normally use for lightdb service
                        without keepalived

1. 启动/停止/重启当前集群

lightdb_service.py -c start/stop/restart

2. 启动/停止/重启所有CN节点

lightdb_service.py -c start/stop/restart --cn_only 

3. 启动/停止/重启所有DN节点

lightdb_service.py -c start/stop/restart --dn_only 

4. 启动/停止/重启所有主节点

lightdb_service.py -c start/stop/restart --primary_only 

5. 启动/停止/重启所有备节点

lightdb_service.py -c start/stop/restart --standby_only 

6. 启动/停止/重启所有DN的主节点, 类似的可以自由组合主备选项和分布式cn,dn 选项

lightdb_service.py -c start/stop/restart --cn_only --primary_only

7. 在启停分布式集群时，可以通过指定--parallel对DN节点进行并发启停

lightdb_service.py -c start/stop/restart --parallel 4 

8. 忽略keepalived

lightdb_service.py -c start/stop/restart --ignore-keepalived

9. 指定节点启停

lightdb_service.py -c start/stop/restart -n 10.20.148.122:54333

三. 异常及主要事项

• 指定主备停止，需要所有节点都是running状态
• 出错后需人工介入
• 脚本使用ssh, 需要配置ssh免密(对本机也需配置免密)
• 对高可用集群默认会操作keepalived, 需要配置sudo 免密，若不需要操作keepalived, 使用上述–ignore-keepalived 选项

错误码

Error Code	Reason
100	Start lightdb failed(please refer to lightdb log for details), not primary
101	Stop lightdb failed(please refer to lightdb log for details), not primary
102	Start primary lightdb failed(please refer to lightdb log for details)
103	Stop primary lightdb failed(please refer to lightdb log for details)
104	The node to operate is not in current lightdb service. used with ‘-n’
105	Get lightdb role failed. 获取lightdb 主备状态失败
106	Cannot get lightdb process status, execute command that will check lightdb port failed. Usually, ssh encryption is not configured. 执行检测lightdb端口的命令失败
107	LightDB process is running, but can’t connect to execute sql.
108	LightDB is stopped, but it should be running.
109	LightDB is running, but it should be stopped.
110	Cannot get ltcluster process status, because execute get command failed. Usually, ssh encryption is not configured or lightdb is stopped
111	ltclusterd is stopped, but it should be running.
112	ltclusterd is running, but it should be stopped.
113	All ltclusterd in cluster are unpaused. but should be paused
114	All ltclusterd in cluster are paused. but should be unpaused
115	All ltclusterd pause status in cluster are unknown.
116	ltclusterd pause status in cluster are inconsistent.
117	ltclusterd start failed. please refer to log for details.
118	ltclusterd stop failed. please refer to log for details.
119	Pause ltclusterd in cluster.
120	Unpause ltclusterd in cluster failed.
121	Cannot get keepalived process status, because execute get command failed.
122	keepalived is stopped, but should be running.
123	keepalived is running, but should be stopped.
124	keepalived start failed.
125	keepalived stop failed.
126	Use for execute command failed in Common scene, now only used for get standby.signal file num error.
127	standby.signal file’s num is wrong when judge lightdb role.
128	Cannot get high available status, because execute command failed.
129	High available status is not right(right meaning all is running correctly).
130	install file about lightdb service’s formart is error.
131	install file about lightdb service’s content is error.