GreenPlum 数据库启动关闭及数据库状态检查

本篇文章主要记录GreenPlum数据库的启动、关闭及状态查询。GreenPlum数据库提供gpstart和gpstop脚本来启动和关闭数据库，可以通过—help参数来查看这些脚本的帮助信息。

view source
 
01 [gpadmin@mdw master]$ gpstart --help
02 COMMAND NAME: gpstart
03 Starts a Greenplum Database system.
04 *****************************************************
05 SYNOPSIS
06 *****************************************************
07 gpstart [-d <master_data_directory>] [-B <parallel_processes>]
08         [-R] [-m] [-y] [-a] [-t <timeout_seconds>]
09         [-l logfile_directory] [-v | -q]
10 gpstart -? | -h | --help
11 gpstart --version
12 *****************************************************
13 DESCRIPTION
14 *****************************************************
15 The gpstart utility is used to start the Greenplum Database server
16 processes. When you start a Greenplum Database system, you are
17 actually starting several postgres database server listener processes
18 at once (the master and all of the segment instances). The gpstart utility
19 handles the startup of the individual instances. Each instance is started
20 in parallel.
21 The first time an administrator runs gpstart, the utility creates a hosts
22 cache file named .gphostcache in the user's home directory. Subsequently,
23 the utility uses this list of hosts to start the system more efficiently.
24 If new hosts are added to the system, you must manually remove this file
25 from the gpadmin user's home directory. The utility will create a new hosts
26 cache file at the next startup.
27 Before you can start a Greenplum Database system, you must have initialized
28 the system using gpinitsystem first.
29 *****************************************************
30 OPTIONS
31 *****************************************************
32 -a (do not prompt)
33   Do not prompt the user for confirmation.
34 -B <parallel_processes>
35   The number of segments to start in parallel. If not specified,
36   the utility will start up to 64 parallel processes depending on
37   how many segment instances it needs to start.
38 -d <master_data_directory>
39   Optional. The master host data directory. If not specified,
40   the value set for $MASTER_DATA_DIRECTORY will be used.
41 -l <logfile_directory>
42   The directory to write the log file. Defaults to ~/gpAdminLogs.
43 -m (master only)
44   Optional. Starts the master instance only, which may be necessary
45   for maintenance tasks. This mode only allows connections to the master
46   in utility mode. For example:
47   PGOPTIONS='-c gp_session_role=utility' psql
48   Note that starting the system in master-only mode is only advisable
49   under supervision of Greenplum support.  Improper use of this option
50   may lead to a split-brain condition and possible data loss.
51 -q (no screen output)
52   Run in quiet mode. Command output is not displayed on the screen,
53   but is still written to the log file.
54 -R (restricted mode)
55   Starts Greenplum Database in restricted mode (only database superusers
56   are allowed to connect).
57 -t | --timeout <number_of_seconds>
58   Specifies a timeout in seconds to wait for a segment instance to
59   start up. If a segment instance was shutdown abnormally (due to
60   power failure or killing its postgres database listener process,
61   for example), it may take longer to start up due to the database
62   recovery and validation process. If not specified, the default timeout
63   is 60 seconds.
64 -v (verbose output)
65   Displays detailed status, progress and error messages output by the utility.
66 -y (do not start standby master)
67   Optional. Do not start the standby master host. The default is to start
68   the standby master host and synchronization process.
69 -? | -h | --help (help)
70   Displays the online help.
71 --version (show utility version)
72   Displays the version of this utility.
73 *****************************************************
74 EXAMPLES
75 *****************************************************
76 Start a Greenplum Database system:
77 gpstart
78 Start a Greenplum Database system in restricted mode
79 (only allow superuser connections):
80 gpstart -R
81 Start the Greenplum master instance only and connect in utility mode:
82 gpstart -m
83 PGOPTIONS='-c gp_session_role=utility' psql
84 Display the online help for the gpstart utility:
85 gpstart -?
86 *****************************************************
87 SEE ALSO
88 *****************************************************
89 gpinitsystem, gpstop

常用的启动参数有以下几个参数：

-a，该模式不需要在启动过程中输入Y进行确认，将直接启动数据库。

-m，只启动Master节点，不启动Segment节点，通常在维护的时候使用。

-y，只启动Master的primary节点，不启动standby节点。

如下为GreenPlum数据库的启动过程。

view source
 
01 [gpadmin@mdw gpdb]$ gpstart -a
02 20160114:10:56:18:013983 gpstart:mdw:gpadmin-[INFO]:-Starting gpstart with args: -a
03 20160114:10:56:18:013983 gpstart:mdw:gpadmin-[INFO]:-Gathering information and validating the environment...
04 20160114:10:56:18:013983 gpstart:mdw:gpadmin-[INFO]:-Greenplum Binary Version: 'postgres (Greenplum Database) 4.3.6.2 build 1'
05 20160114:10:56:18:013983 gpstart:mdw:gpadmin-[INFO]:-Greenplum Catalog Version: '201310150'
06 20160114:10:56:18:013983 gpstart:mdw:gpadmin-[INFO]:-Starting Master instance in admin mode
07 20160114:10:56:19:013983 gpstart:mdw:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
08 20160114:10:56:19:013983 gpstart:mdw:gpadmin-[INFO]:-Obtaining Segment details from master...
09 20160114:10:56:19:013983 gpstart:mdw:gpadmin-[INFO]:-Setting new master era
10 20160114:10:56:19:013983 gpstart:mdw:gpadmin-[INFO]:-Master Started...
11 20160114:10:56:19:013983 gpstart:mdw:gpadmin-[INFO]:-Shutting down master
12 20160114:10:56:21:013983 gpstart:mdw:gpadmin-[INFO]:-Commencing parallel segment instance startup, please wait
13 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-Process results...
14 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-----------------------------------------------------
15 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-   Successful segment starts      = 3
16 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-   Failed segment starts          = 0
17 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-   Skipped segment starts (segments are marked down in configuration)   = 0
18 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-----------------------------------------------------
19 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-
20 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-Successfully started 3 of 3 segment instances
21 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-----------------------------------------------------
22 20160114:10:56:23:013983 gpstart:mdw:gpadmin-[INFO]:-Starting Master instance mdw directory /gpdb/gpdata/master/gpseg-1
23 20160114:10:56:24:013983 gpstart:mdw:gpadmin-[INFO]:-Command pg_ctl reports Master mdw instance active
24 20160114:10:56:24:013983 gpstart:mdw:gpadmin-[INFO]:-No standby master configured.  skipping...
25 20160114:10:56:24:013983 gpstart:mdw:gpadmin-[INFO]:-Database successfully started

可以通过gpstate脚本来查看数据库的状态信息，该命令的帮助信息如下：

view source
 
001 [gpadmin@mdw master]$ gpstate --help
002 COMMAND NAME: gpstate
003 Shows the status of a running Greenplum Database system.
004 *****************************************************
005 SYNOPSIS
006 *****************************************************
007 gpstate [-d <master_data_directory>] [-B <parallel_processes>]
008         [-s | -b | -Q | -e] [-m | -c] [-p] [-i] [-f]
009         [-v | -q] [-l <log_directory>]
010 gpstate -? | -h | --help
011 *****************************************************
012 DESCRIPTION
013 *****************************************************
014 The gpstate utility displays information about a running
015 Greenplum Database instance. There is additional information
016 you may want to know about a Greenplum Database system, since
017 it is comprised of multiple PostgreSQL database instances (segments)
018 spanning multiple machines. The gpstate utility provides
019 additional status information for a Greenplum Database system,
020 such as:
021 * Which segments are down.
022 * Master and segment configuration information (hosts,
023   data directories, etc.).
024 * The ports used by the system.
025 * A mapping of primary segments to their corresponding
026   mirror segments.
027 *****************************************************
028 OPTIONS
029 *****************************************************
030 -b (brief status)
031   Optional. Display a brief summary of the state of the
032   Greenplum Database system. This is the default option.
033 -B <parallel_processes>
034   The number of segments to check in parallel. If not specified,
035   the utility will start up to 60 parallel processes depending on
036   how many segment instances it needs to check.
037 -c (show primary to mirror mappings)
038   Optional. Display mapping of primary segments to their
039   corresponding mirror segments.
040 -d <master_data_directory>
041   Optional. The master data directory. If not specified, the
042   value set for $MASTER_DATA_DIRECTORY will be used.
043 -e (show segments with mirror status issues)
044   Show details on primary/mirror segment pairs that have
045   potential issues such as 1) the active segment is running in
046   change tracking mode, meaning a segment is down 2) the active
047   segment is in resynchronization mode, meaning it is catching
048   up changes to the mirror 3) a segment is not in its preferred role,
049   for example a segment that was a primary at system initialization
050   time is now acting as a mirror, meaning you may have one or more
051   segment hosts with unbalanced processing load.
052 -f (show standby master details)
053   Display details of the standby master host if configured.
054 -i (show Greenplum Database version)
055   Display the Greenplum Database software version information
056   for each instance.
057 -l <logfile_directory>
058   The directory to write the log file. Defaults to ~/gpAdminLogs.
059 -m (list mirrors)
060    Optional. List the mirror segment instances in the system,
061   their current role, and synchronization status.
062 -p (show ports)
063   List the port numbers used throughout the Greenplum Database
064   system.
065 -q (no screen output)
066   Optional. Run in quiet mode. Except for warning messages, command
067   output is not displayed on the screen. However, this information
068   is still written to the log file.
069 -Q (quick status)
070   Optional. Checks segment status in the system catalog on
071   the master host. Does not poll the segments for status.
072 -s (detailed status)
073   Optional. Displays detailed status information for the
074   Greenplum Database system.
075 -v (verbose output)
076   Optional. Displays error messages and outputs detailed status
077   and progress information.
078 -? | -h | --help (help)
079   Displays the online help.
080 *****************************************************
081 OUTPUT DEFINITIONS FOR DETAIL VIEW
082 *****************************************************
083 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
084 MASTER OUTPUT DATA
085 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
086 * Master host - host name of the master
087 * Master postgres process ID - PID of the master postgres database
088                                listener process
089 * Master data directory - file system location of the master data directory
090 * Master port - port of the master database listener process
091 * Master current role - dispatch = regular operating mode
092                         utility = maintenance mode
093 * Greenplum array configuration type - Standard = one NIC per host
094                                        Multi-Home = multiple NICs per host
095 * Greenplum initsystem version - version of Greenplum Database when
096                                  system was first initialized
097 * Greenplum current version - current version of Greenplum Database
098 * Postgres version - version of PostgreSQL that Greenplum Database
099                      is based on
100 * Greenplum mirroring status - physical mirroring, SAN or none
101 * Master standby - host name of the standby master
102 * Standby master state - status of the standby master: active or passive
103 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
104 SEGMENT OUTPUT DATA
105 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
106 * Hostname - system-configured host name
107 * Address - network address host name (NIC name)
108 * Datadir - file system location of segment data directory
109 * Port - port number of segment postgres database listener process
110 * Current Role - current role of a segment: Mirror or Primary
111 * Preferred Role - role at system initialization time: Mirror or Primary
112 * Mirror Status - status of a primary/mirror segment pair:
113                 Synchronized = data is up to date on both
114                 Resynchronizing = data is currently being copied from one to the other
115                 Change Tracking = segment down and active segment is logging changes
116 * Change tracking data size - when in Change Tracking mode, the size of the change
117                             log file (may grow and shrink as compression is applied)
118 * Estimated total data to synchronize - when in Resynchronization mode, the estimated
119                                       size of data left to syncronize
120 * Data synchronized - when in Resynchronization mode, the estimated size of data
121                     that has already been syncronized
122 * Estimated resync progress with mirror - when in Resynchronization mode, the
123                                         estimated percentage of completion
124 * Estimated resync end time - when in Resynchronization mode, the estimated
125                             time to complete
126 * File postmaster.pid - status of postmaster.pid lock file: Found or Missing
127 * PID from postmaster.pid file - PID found in the postmaster.pid file
128 * Lock files in /tmp - a segment port lock file for its postgres process is
129                        created in /tmp (file is removed when a segment shuts down)
130 * Active PID - active process ID of a segment
131 * Master reports status as - segment status as reported in the system catalog:
132                            Up or Down
133 Database status - status of Greenplum Database to incoming requests:
134                 Up, Down, or Suspended. A Suspended state means database
135                 activity is temporarily paused while a segment transitions from
136                 one state to another.
137 *****************************************************
138 EXAMPLES
139 *****************************************************
140 Show detailed status information of a Greenplum Database system:
141    gpstate -s
142 Do a quick check for down segments in the master host system catalog:
143    gpstate -Q
144 Show information about mirror segment instances:
145    gpstate -m
146 Show information about the standby master configuration:
147    gpstate -f
148 Display the Greenplum software version information:
149    gpstate -i
150 *****************************************************
151 SEE ALSO
152 *****************************************************
153

常用的参数如下：

-s，详细信息。

-m，Mirror信息。

-f，Master的Standby信息。

-e，Segment的Mirror信息。

-i，版本信息。

如下为数据库启动状态通过gpstate –s查询的数据库状态信息。

view source
 
01 [gpadmin@mdw gpdb]$ gpstate -s
02 20160114:10:56:40:014082 gpstate:mdw:gpadmin-[INFO]:-Starting gpstate with args: -s
03 20160114:10:56:40:014082 gpstate:mdw:gpadmin-[INFO]:-local Greenplum Version: 'postgres (Greenplum Database) 4.3.6.2 build 1'
04 20160114:10:56:40:014082 gpstate:mdw:gpadmin-[INFO]:-master Greenplum Version: 'PostgreSQL 8.2.15 (Greenplum Database 4.3.6.2 build 1) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Nov 12 2015 23:50:28'
05 20160114:10:56:40:014082 gpstate:mdw:gpadmin-[INFO]:-Obtaining Segment details from master...
06 20160114:10:56:40:014082 gpstate:mdw:gpadmin-[INFO]:-Gathering data from segments...
07 .
08 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-----------------------------------------------------
09 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:--Master Configuration & Status
10 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-----------------------------------------------------
11 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Master host                    = mdw
12 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Master postgres process ID     = 14042
13 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Master data directory          = /gpdb/gpdata/master/gpseg-1
14 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Master port                    = 5432
15 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Master current role            = dispatch
16 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Greenplum initsystem version   = 4.3.6.2 build 1
17 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Greenplum current version      = PostgreSQL 8.2.15 (Greenplum Database 4.3.6.2 build 1) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Nov 12 2015 23:50:28
18 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Postgres version               = 8.2.15
19 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Master standby                 = No master standby configured
20 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-----------------------------------------------------
21 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-Segment Instance Status Report
22 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-----------------------------------------------------
23 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Segment Info
24 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Hostname                          = sdw1
25 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Address                           = sdw1
26 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Datadir                           = /gpdb/gpdata/primary/gpseg0
27 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Port                              = 40000
28 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Status
29 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      PID                               = 17195
30 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Configuration reports status as   = Up
31 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Database status                   = Up
32 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-----------------------------------------------------
33 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Segment Info
34 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Hostname                          = sdw2
35 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Address                           = sdw2
36 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Datadir                           = /gpdb/gpdata/primary/gpseg1
37 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Port                              = 40000
38 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Status
39 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      PID                               = 17111
40 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Configuration reports status as   = Up
41 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Database status                   = Up
42 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-----------------------------------------------------
43 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Segment Info
44 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Hostname                          = sdw3
45 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Address                           = sdw3
46 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Datadir                           = /gpdb/gpdata/primary/gpseg2
47 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Port                              = 40000
48 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-   Status
49 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      PID                               = 23033
50 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Configuration reports status as   = Up
51 20160114:10:56:41:014082 gpstate:mdw:gpadmin-[INFO]:-      Database status                   = Up

gpstop脚本用来关闭、重启、加载新修改的参数以至生效等功能，下面是gpstop命令的帮助信息。

view source
 
001 [gpadmin@mdw master]$ gpstop --help
002 COMMAND NAME: gpstop
003 Stops or restarts a Greenplum Database system.
004 *****************************************************
005 SYNOPSIS
006 *****************************************************
007 gpstop [-d <master_data_directory>] [-B <parallel_processes>]
008        [-M smart | fast | immediate] [-t <timeout_seconds>]
009        [-r] [-y] [-a] [-l <logfile_directory>] [-v | -q]
010 gpstop -m [-d <master_data_directory>] [-y] [-l <logfile_directory>]
011        [-v | -q]
012 gpstop -u [-d <master_data_directory>] [-l <logfile_directory>]
013           [-v | -q]
014 gpstop --version
015 gpstop -? | -h | --help
016 *****************************************************
017 DESCRIPTION
018 *****************************************************
019 The gpstop utility is used to stop the database servers that
020 comprise a Greenplum Database system. When you stop a Greenplum
021 Database system, you are actually stopping several postgres
022 database server processes at once (the master and all of the
023 segment instances). The gpstop utility handles the shutdown
024 of the individual instances. Each instance is shutdown in parallel.
025 By default, you are not allowed to shut down Greenplum Database
026 if there are any client connections to the database. Use
027 the -M fast option to roll back and terminate any connections
028 before shutting down. If there are any transactions in progress,
029 the default behavior is to wait for them to commit before
030 shutting down. Use the -M fast option to roll back open
031 transactions.
032 With the -u option, the utility uploads changes made to the
033 master pg_hba.conf file or to runtime configuration parameters
034 in the master postgresql.conf file without interruption of
035 service. Note that any active sessions will not pickup the
036 changes until they reconnect to the database.
037 *****************************************************
038 OPTIONS
039 *****************************************************
040 -a (do not prompt)
041  Do not prompt the user for confirmation.
042 -B <parallel_processes>
043  The number of segments to stop in parallel. If not specified,
044  the utility will start up to 64 parallel processes depending
045  on how many segment instances it needs to stop.
046 -d <master_data_directory>
047  Optional. The master host data directory. If not specified,
048  the value set for $MASTER_DATA_DIRECTORY will be used.
049 -l <logfile_directory>
050  The directory to write the log file. Defaults to ~/gpAdminLogs.
051 -m (master only)
052  Optional. Shuts down a Greenplum master instance that was
053  started in maintenance mode.
054 -M fast (fast shutdown - rollback)
055  Fast shut down. Any transactions in progress are interrupted  and rolled back.
056 -M immediate (immediate shutdown - abort)
057  Immediate shut down. Any transactions in progress are aborted. This
058  shutdown mode is not recommended, and in some circumstances can cause
059  database corruption requiring manual recovery.
060  This mode kills all postgres processes without allowing the database
061  server to complete transaction processing or clean up any temporary or
062  in-process work files.
063 -M smart (smart shutdown - warn)
064   Smart shut down. If there are active connections, this command
065  fails with a warning. This is the default shutdown mode.
066 -q (no screen output)
067  Run in quiet mode. Command output is not displayed on the
068  screen, but is still written to the log file.
069 -r (restart)
070  Restart after shutdown is complete.
071 -t <timeout_seconds>
072  Specifies a timeout threshold (in seconds) to wait for a
073  segment instance to shutdown. If a segment instance does not
074  shutdown in the specified number of seconds, gpstop displays
075  a message indicating that one or more segments are still in
076  the process of shutting down and that you cannot restart
077  Greenplum Database until the segment instance(s) are stopped.
078  This option is useful in situations where gpstop is executed
079  and there are very large transactions that need to rollback.
080  These large transactions can take over a minute to rollback
081  and surpass the default timeout period of 600 seconds.
082 -u (reload pg_hba.conf and postgresql.conf files only)
083  This option reloads the pg_hba.conf files of the master and
084  segments and the runtime parameters of the postgresql.conf files
085  but does not shutdown the Greenplum Database array. Use this
086  option to make new configuration settings active after editing
087  postgresql.conf or pg_hba.conf. Note that this only applies to
088  configuration parameters that are designated as runtime parameters.
089 -v (verbose output)
090  Displays detailed status, progress and error messages output
091  by the utility.
092 --version (show utility version)
093  Displays the version of this utility.
094 -y (do not stop standby master)
095  Do not stop the standby master process. The default is to stop  the standby master.
096 -? | -h | --help (help)
097  Displays the online help.
098 *****************************************************
099 EXAMPLES
100 *****************************************************
101 Stop a Greenplum Database system in smart mode:
102   gpstop
103 Stop a Greenplum Database system in fast mode:
104   gpstop -M fast
105 Stop all segment instances and then restart the system:
106   gpstop -r
107 Stop a master instance that was started in maintenance mode:
108   gpstop -m
109 Reload the postgresql.conf and pg_hba.conf files after
110 making runtime configuration parameter changes but do not
111 shutdown the Greenplum Database array:
112   gpstop -u
113 *****************************************************
114 SEE ALSO
115 *****************************************************
116 gpstart

常用的参数如下：

-a，不需要输入Y确认是否关闭，将直接关闭数据库。

-m，只关闭Master节点。

-r，重启数据库。

-u，加载参数文件，使修改的参数生效。

-M，设置关闭数据库的级别，有三种级别，fast、immediate和smart，gpstop –M fast，所有连接将中断并且回滚。gpstop –M immediate，强制关闭数据库，这种方式是不一致的关闭模式，不建议使用。Immediate smart，这是默认的关闭级别，所有连接的会话会收到关闭警告，不允许新链接访问数据库。下面是默认级别关闭GreenPlum数据库的相关信息。

view source
 
01 [gpadmin@mdw gpdb]$ gpstop -a
02 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-Starting gpstop with args: -a
03 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-Gathering information and validating the environment...
04 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
05 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-Obtaining Segment details from master...
06 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-Greenplum Version: 'postgres (Greenplum Database) 4.3.6.2 build 1'
07 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-There are 0 connections to the database
08 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-Commencing Master instance shutdown with mode='smart'
09 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-Master host=mdw
10 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-Commencing Master instance shutdown with mode=smart
11 20160114:10:55:17:013853 gpstop:mdw:gpadmin-[INFO]:-Master segment instance directory=/gpdb/gpdata/master/gpseg-1
12 20160114:10:55:18:013853 gpstop:mdw:gpadmin-[INFO]:-Attempting forceful termination of any leftover master process
13 20160114:10:55:18:013853 gpstop:mdw:gpadmin-[INFO]:-Terminating processes for segment /gpdb/gpdata/master/gpseg-1
14 20160114:10:55:18:013853 gpstop:mdw:gpadmin-[INFO]:-No standby master host configured
15 20160114:10:55:18:013853 gpstop:mdw:gpadmin-[INFO]:-Commencing parallel segment instance shutdown, please wait...
16 20160114:10:55:18:013853 gpstop:mdw:gpadmin-[INFO]:-0.00% of jobs completed
17 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-100.00% of jobs completed
18 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-----------------------------------------------------
19 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-   Segments stopped successfully      = 3
20 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-   Segments with errors during stop   = 0
21 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-----------------------------------------------------
22 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-Successfully shutdown 3 of 3 segment instances
23 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-Database successfully shutdown with no errors reported
24 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-Cleaning up leftover gpmmon process
25 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-No leftover gpmmon process found
26 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-Cleaning up leftover gpsmon processes
27 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-No leftover gpsmon processes on some hosts. not attempting forceful termination on these hosts
28 20160114:10:55:28:013853 gpstop:mdw:gpadmin-[INFO]:-Cleaning up leftover shared memory

本文固定链接: http://www.dbdream.com.cn/2016/01/greenplum%e6%95%b0%e6%8d%ae%e5%ba%93%e5%90%af%e5%8a%a8%e5%85%b3%e9%97%ad%e5%8f%8a%e6%95%b0%e6%8d%ae%e5%ba%93%e7%8a%b6%e6%80%81%e6%a3%80%e6%9f%a5/ | 信春哥，系统稳，闭眼上线不回滚！

【推广】 免费学中医，健康全家人
原文地址：https://www.cnblogs.com/xibuhaohao/p/11131117.html