CNDBA社区
1 理论说明
在前文我们了解了单机版OpenGauss的安装，如下：
Linux 7.6 平台 openGauss 3.1.1 单机版 安装手册
https://www.cndba.cn/dave/article/116448
本文我们了解一下OpenGauss的启停操作。
OpenGauss的启停有个工具：gs_om 和 gs_ctl。
- gs_om 是服务端工具，该工具提供对openGauss进行维护，包括启动、停止、查询openGauss状态、查询静态配置、生成静态配置文件、查询openGauss状态详细信息、生成动态配置文件、SSL证书替换、显示帮助信息和显示版本号信息等功能。同时注意改命令需要使用omm用户来执行。
- gs_ctl 是系统内部使用的工具，通过gs_ctl是也可以用来启停数据库服务和查询数据库状态。主要供openGauss管理模块调用。gs_ctl工具也要由omm用户执行。
该工具可以：
启动、停止、重启openGauss节点。
在不停止数据库的情况下，重新加载配置文件（postgresql.conf，pg_hba.conf）。
主备切换、主备状态查询、重建和重建状态查询。
其他工具参考之前的博客：
openGauss 数据库 工具 清单
https://www.cndba.cn/dave/article/116476
2 操作示例
2.1 查看帮助
[dave@www.cndba.cn ~]# ps -ef|grep openGauss
root 19367 17786 0 20:40 pts/7 00:00:00 grep --color=auto openGauss
omm 43976 1 7 Feb19 ? 1-07:46:39 /data/openGauss/install/app/bin/gaussdb -D /data/openGauss/install/data/dn
[dave@www.cndba.cn ~]# su - omm
Last login: Sun Feb 19 16:36:54 CST 2023 on pts/9
[dave@www.cndba.cn ~]$ gs_om -V
gs_om (openGauss OM 3.0.3 build 43c77d67) compiled at 2023-01-10 22:52:03 commit 0 last mr
[dave@www.cndba.cn ~]$ gs_om -?
gs_om is a utility to manage a cluster.
Usage:
gs_om -? | --help
gs_om -V | --version
OLAP scene:
gs_om -t start [-h HOSTNAME] [-D dataDir] [--time-out=SECS]
[--security-mode=MODE] [--cluster-number=None] [-l LOGFILE]
gs_om -t stop [-h HOSTNAME] [-D dataDir] [--time-out=SECS] [-m MODE]
[-l LOGFILE]
gs_om -t restart [-h HOSTNAME] [-D dataDir] [--time-out=SECS]
[--security-mode=MODE] [-l LOGFILE] [-m MODE]
gs_om -t status [-h HOSTNAME] [-o OUTPUT] [--detail] [--all] [--az=AZ] [-l LOGFILE]
[--time-out=SECS]
gs_om -t generateconf -X XMLFILE [--distribute] [-l LOGFILE]
gs_om -t generateconf --old-values=old --new-values=new [--distribute] [-l LOGFILE]
gs_om -t cert [--cert-file=CERTFILE | --rollback] [-L] [-l LOGFILE]
gs_om -t kerberos -m [install|uninstall] -U USER [-l LOGFILE]
[--krb-server|--krb-client]
gs_om -t view [-o OUTPUT]
gs_om -t query [-o OUTPUT] [--time-out=SECS]
gs_om -t refreshconf
General options:
-t Type of the OM command.
-l Path of log file.
-?, --help Show help information for this utility,
and exit the command line mode.
-V, --version Show version information.
Options for start
-h Name of the host to be started.
-D Path of dn
--time-out=SECS Maximum waiting time when start the
cluster or node.
--security-mode=MODE database start with security mode: on or off
on: start with security mode
off: start without security mode
--cluster-number database start with kernel version corresponding
to version.cfg
Options for stop
-h Name of the host to be shut down.
-m, --mode=MODE Shutdown mode. It can be f (fast),
or i (immediate).
-D Path of dn
--time-out=SECS Maximum waiting time when start the cluster
or node.
Options for restart
-h Name of the host to be started.
-m, --mode=MODE Shutdown mode. It can be f (fast),
or i (immediate).
-D Path of dn
--time-out=SECS Maximum waiting time when start the
cluster or node.
--security-mode=MODE database start with security mode: on or off
on: start with security mode
off: start without security mode
Options for status
-h Name of the host whose status is to be
queried.
--az=AZ Name of the single az whose status is to
be queried.
-o Save the result to the specified file.
--detail Show detailed status information.
--all Show all database node status information.
--time-out=SECS Maximum waiting time when query cluster status.
Options for generating configuration files
-X Path of the XML configuration file.
--old-values IP or hostname or port like "host1,ip1,port1"
--new-values IP or hostname or port like "host1,ip1,port1"
--distribute Distribute the static configuration file
to installation directory of cluster nodes.
Options for cert
--cert-file Path of cert file.
--rollback Perform rollback SSL cert files.
-L local mode.
Options for kerberos
-m Kerberos management mode. It can be
install or uninstall.
-U %s cluster user.
Install options:
--krb-server Execute install for server. This parameter
only work for install
--krb-client Execute install for client. This parameter
only work for install
[dave@www.cndba.cn ~]$
[dave@www.cndba.cn ~]$ gs_ctl --help
gs_ctl is a utility to initialize, start, stop, or control a openGauss server.
Usage:
gs_ctl init[db] [-D DATADIR] [-s] [-o "OPTIONS"]
gs_ctl start [-w] [-t SECS] [-D DATADIR] [-s] [-l FILENAME] [-o "OPTIONS"] [-M SERVERMODE]
gs_ctl restart [-w] [-t SECS] [-D DATADIR] [-s] [-m SHUTDOWN-MODE]
[-o "OPTIONS"]
gs_ctl build [-D DATADIR] [-b MODE] [-r SECS] [-C CONNECTOR] [-q] [-M SERVERMODE]
gs_ctl stop [-W] [-t SECS] [-D DATADIR] [-s] [-m SHUTDOWN-MODE]
gs_ctl reload [-D DATADIR] [-s]
gs_ctl status [-D DATADIR]
gs_ctl finishredo [-D DATADIR] [-s]
gs_ctl failover [-W] [-t SECS] [-D DATADIR] [-U USERNAME] [-P PASSWORD] [-T TERM]
gs_ctl switchover [-W] [-D DATADIR] [-m SWITCHOVER-MODE] [-U USERNAME] [-P PASSWORD] [-f]
gs_ctl query [-D DATADIR] [-U USERNAME] [-P PASSWORD] [-L lsn]
gs_ctl notify -M SERVERMODE [-D DATADIR] [-U USERNAME] [-P PASSWORD]
gs_ctl kill SIGNALNAME PID
gs_ctl querybuild [-D DATADIR]
gs_ctl copy [-D DATADIR] [-Q COPYMODE]
gs_ctl stack [-D DATADIR] [-I lwtid]
gs_ctl member [-O OPERATION] [-u DCF-NODE-ID] [-i DCF-NODE-IP] [-e DCF-NODE-PORT] [-D PRIMARY-DATADIR]
gs_ctl changerole [-R DCF-ROLE] [-D DATADIR] [-t SEC]
gs_ctl setrunmode [-x XMODE] [-v VOTE-NUM] [-D PRIMARY-DATADIR]
Common options:
-b, --mode=MODE the mode of building the datanode or coordinator.MODE can be "full", "incremental", "auto", "standby_full", "copy_secure_files", "copy_upgrade_file", "cross_cluster_full", "cross_cluster_incremental", "cross_cluster_standby_full"
-D, --pgdata=DATADIR location of the database storage area
-s, --silent only print errors, no informational messages
-t, --timeout=SECS seconds to wait when using -w option
-V, --version output version information, then exit
-w wait until operation completes
-W do not wait until operation completes
-M the database start as the appointed mode
-T Failover requires a term
-q do not start automatically after build finishing, needed start by caller
-d more debug info will be print
-L query lsn:XX/XX validity and show the max_lsn
-P PASSWORD password of account to connect local server
-U USERNAME user name of account to connect local server
-Z NODE-TYPE can be "single_node"
-?, -h, --help show this help, then exit
(The default is to wait for shutdown, start and restart.)
If the -D option is omitted, the environment variable PGDATA is used.
Options for start or restart:
-c, --core-files allow openGauss to produce core files
-l, --log=FILENAME write (or append) server log to FILENAME
-o OPTIONS command line options to pass to openGauss
(openGauss server executable) or gs_initdb
-p PATH-TO-POSTGRES normally not necessary
Options for stop or restart:
-m, --mode=MODE MODE can be "fast" or "immediate"
Options for xlog copy:
-Q, --mode=MODE MODE can be "copy_from_local", "force_copy_from_local", "copy_from_share"
Options for DCF:
-O, --operation=OPERATION Operation of adding or removing or change a DN configuration.
-u, --nodeid=DCF-NODE-ID It is required for member command.
-i, --ip=DCF-NODE-IP It is required when OPERATION of member command is "add".
-e, --port=DCF-NODE-PORT It is required when OPERATION of member command is "add".
-R, --role=DCF-NODE-ROLE The option is "follower" or "passive".
-v, --votenum=VOTE-NUM It is required when XMODE is "minority".
-x, --xmode=XMODE The option can be "minority" or "normal" mode in DCF.
-G, The option is a int type to set group number in DCF.
--priority, The option is a int type to set priority number in DCF.
OPERATION are:
add add a member to DCF configuration.
remove remove a member from DCF configuration.
change change DCF node configuration.
XMODE are:
minority the leader in DCF can reach consensus when getting less than half nodes' response.
normal the leader in DCF can reach consensus when getting more than half nodes' response.
DCF-NODE-ROLE are:
follower follower role in DCF
passive passive role in DCF
Options for stack:
-I lwtid light weight thread id. get stack of specific thread according to it's lwtid
Shutdown modes are:
fast quit directly, with proper shutdown
immediate quit without complete shutdown; will lead to recovery on restart
Switchover modes are:
-f quit directly, with proper shutdown and do not perform checkpoint
fast demote primary directly, with proper shutdown
SERVERMODE are:
primary database system run as a primary server, send xlog to standby server
standby database system run as a standby server, receive xlog from primary server
cascade_standby database system run as a cascade standby server, receive xlog from standby server
pending database system run as a pending server, wait for promoting to primary or demoting to standby
Only used in start command
Allowed signal names for kill:
ABRT HUP INT QUIT TERM USR1 USR2
Build connection option:
-r, --recvtimeout=INTERVAL time that receiver waits for communication from server (in seconds)
-C, connector CN/DN connect to specified CN/DN for build
Report bugs to openGauss community by raising an issue.
[dave@www.cndba.cn ~]$
2.2 查看状态
[dave@www.cndba.cn ~]$ gs_om -t status
Password:
-----------------------------------------------------------------------
cluster_name : dbCluster
cluster_state : Normal
redistributing : No
-----------------------------------------------------------------------
[dave@www.cndba.cn ~]$ gs_om -t view
NodeHeader:
version:301
time:1676795851
nodeCount:1
node:1
============================================================
azName:AZ1
azPriority:1
node :1
nodeName:hf-172-31-101-18
ssh channel :
sshChannel 1:172.31.101.18
datanodeCount :1
datanodeInstanceType :primary
datanode 1:
datanodeLocalDataPath :/data/openGauss/install/data/dn
datanodeXlogPath :
datanodeListenIP 1:172.31.101.18
datanodePort :15400
datanodeLocalHAIP 1:172.31.101.18
datanodeLocalHAPort :15401
dn_replication_num: 1
============================================================
[dave@www.cndba.cn ~]$ ps -ef|grep openGauss
omm 25729 19387 0 20:51 pts/7 00:00:00 grep --color=auto openGauss
omm 43976 1 7 Feb19 ? 1-07:47:31 /data/openGauss/install/app/bin/gaussdb -D /data/openGauss/install/data/dn
[dave@www.cndba.cn ~]$ gs_om -t query
Password:
[ Cluster State ]
cluster_state : Normal
redistributing : No
current_az : AZ_ALL
[ Datanode State ]
node node_ip port instance state
--------------------------------------------------------------------------------
1 hf-172-31-101-18 172.31.101.18 15400 6001 P Primary Normal
[dave@www.cndba.cn ~]$
2.3 启停实例
2.3.1 gs_ctl 命令
[dave@www.cndba.cn ~]$ ps -ef|grep openGauss
omm 31231 19387 0 21:04 pts/7 00:00:00 grep --color=auto openGauss
omm 43976 1 7 Feb19 ? 1-07:48:25 /data/openGauss/install/app/bin/gaussdb -D /data/openGauss/install/data/dn
[dave@www.cndba.cn ~]$ gs_ctl stop -D /data/openGauss/install/data/dn
[dave@www.cndba.cn ~]$ gs_ctl start -D /data/openGauss/install/data/dn
[dave@www.cndba.cn ~]$ gs_ctl restart -D /data/openGauss/install/data/dn
[dave@www.cndba.cn ~]$ gs_ctl status -D /data/openGauss/install/data/dn
[2023-03-09 21:08:48.214][33319][][gs_ctl]: gs_ctl status,datadir is /data/openGauss/install/data/dn
gs_ctl: server is running (PID: 31773)
/data/openGauss/install/app/bin/gaussdb "-D" "/data/openGauss/install/data/dn"
[dave@www.cndba.cn ~]$
2.3.2 gs_om 命令
[dave@www.cndba.cn ~]$ gs_om -t stop -D /data/openGauss/install/data/dn
Stopping cluster.
=========================================
Successfully stopped cluster.
=========================================
End stop cluster.
[dave@www.cndba.cn ~]$ gs_om -t start -D /data/openGauss/install/data/dn
Starting cluster.
=========================================
[SUCCESS] hf-172-31-101-18
2023-03-09 21:10:01.833 6409daa9.1 [unknown] 140487301625536 [unknown] 0 dn_6001 01000 0 [BACKEND] WARNING: could not create any HA TCP/IP sockets
2023-03-09 21:10:01.833 6409daa9.1 [unknown] 140487301625536 [unknown] 0 dn_6001 01000 0 [BACKEND] WARNING: could not create any HA TCP/IP sockets
2023-03-09 21:10:01.834 6409daa9.1 [unknown] 140487301625536 [unknown] 0 dn_6001 01000 0 [BACKEND] WARNING: No explicit IP is configured for listen_addresses GUC.
Password:
=========================================
Successfully started.
[dave@www.cndba.cn ~]$ gs_om -t restart -D /data/openGauss/install/data/dn
Stopping cluster.
=========================================
Successfully stopped cluster.
=========================================
End stop cluster.
Starting cluster.
=========================================
[SUCCESS] hf-172-31-101-18
2023-03-09 21:10:24.197 6409dac0.1 [unknown] 139829065660096 [unknown] 0 dn_6001 01000 0 [BACKEND] WARNING: could not create any HA TCP/IP sockets
2023-03-09 21:10:24.197 6409dac0.1 [unknown] 139829065660096 [unknown] 0 dn_6001 01000 0 [BACKEND] WARNING: could not create any HA TCP/IP sockets
2023-03-09 21:10:24.198 6409dac0.1 [unknown] 139829065660096 [unknown] 0 dn_6001 01000 0 [BACKEND] WARNING: No explicit IP is configured for listen_addresses GUC.
Password:
=========================================
Successfully started.
[dave@www.cndba.cn ~]$
[dave@www.cndba.cn ~]$ gs_om -t status
Password:
-----------------------------------------------------------------------
cluster_name : dbCluster
cluster_state : Normal
redistributing : No
-----------------------------------------------------------------------
[dave@www.cndba.cn ~]$
两个工具的对比：
gs_om 作为服务端的工具，输出的日志更简洁一些，而gs_ctl 作为数据库内部工具，会显示数据库内部的实际的信息，所以内容更多，也更底层一些。
版权声明：本文为博主原创文章，未经博主允许不得转载。
