Deepgreen/Greenplum删除节点步骤
Greenplum和Deepgreen官方都没有给出删除节点的方法和建议，但实际上，我们可以对节点进行删除。由于不确定性，删除节点极有可能导致其他的问题，所以还行做好备份，谨慎而为。下面是具体的步骤：
1.查看数据库当前状态（12个实例）
[gpadmin@sdw1 ~]$ gpstate
20170816:12:53:25:097578 gpstate:sdw1:gpadmin-[INFO]:-Starting gpstate with args:
20170816:12:53:25:097578 gpstate:sdw1:gpadmin-[INFO]:-local Greenplum Version: 'postgres (Greenplum Database) 4.3.99.00 build Deepgreen DB'
20170816:12:53:25:097578 gpstate:sdw1:gpadmin-[INFO]:-master Greenplum Version: 'PostgreSQL 8.2.15 (Greenplum Database 4.3.99.00 build Deepgreen DB) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.9.2 20150212 (Red Hat 4.9.2-6) compiled on Jul 6 2017 03:04:10'
20170816:12:53:25:097578 gpstate:sdw1:gpadmin-[INFO]:-Obtaining Segment details from master...
20170816:12:53:25:097578 gpstate:sdw1:gpadmin-[INFO]:-Gathering data from segments...
..
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:-Greenplum instance status summary
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Master instance = Active
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Master standby = No master standby configured
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total segment instance count from metadata = 12
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Primary Segment Status
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total primary segments = 12
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total primary segment valid (at master) = 12
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total primary segment failures (at master) = 0
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total number of postmaster.pid files missing = 0
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total number of postmaster.pid files found = 12
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total number of postmaster.pid PIDs missing = 0
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total number of postmaster.pid PIDs found = 12
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total number of /tmp lock files missing = 0
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total number of /tmp lock files found = 12
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total number postmaster processes missing = 0
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Total number postmaster processes found = 12
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Mirror Segment Status
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:- Mirrors not configured on this array
20170816:12:53:27:097578 gpstate:sdw1:gpadmin-[INFO]:-----------------------------------------------------
2.并行备份数据库
使用 gpcrondump 命令备份数据库，这里不赘述，不明白的可以翻看文档。
3.关闭当前数据库
[gpadmin@sdw1 ~]$ gpstop -M fast
20170816:12:54:10:097793 gpstop:sdw1:gpadmin-[INFO]:-Starting gpstop with args: -M fast
20170816:12:54:10:097793 gpstop:sdw1:gpadmin-[INFO]:-Gathering information and validating the environment...
20170816:12:54:10:097793 gpstop:sdw1:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
20170816:12:54:10:097793 gpstop:sdw1:gpadmin-[INFO]:-Obtaining Segment details from master...
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:-Greenplum Version: 'postgres (Greenplum Database) 4.3.99.00 build Deepgreen DB'
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:---------------------------------------------
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:-Master instance parameters
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:---------------------------------------------
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- Master Greenplum instance process active PID = 31250
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- Database = template1
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- Master port = 5432
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- Master directory = /hgdata/master/hgdwseg-1
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- Shutdown mode = fast
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- Timeout = 120
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- Shutdown Master standby host = Off
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:---------------------------------------------
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:-Segment instances that will be shutdown:
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:---------------------------------------------
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- Host Datadir Port Status
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg0 25432 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg1 25433 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg2 25434 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg3 25435 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg4 25436 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg5 25437 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg6 25438 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg7 25439 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg8 25440 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg9 25441 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg10 25442 u
20170816:12:54:11:097793 gpstop:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg11 25443 u
Continue with Greenplum instance shutdown Yy|Nn (default=N):
> y
20170816:12:54:12:097793 gpstop:sdw1:gpadmin-[INFO]:-There are 0 connections to the database
20170816:12:54:12:097793 gpstop:sdw1:gpadmin-[INFO]:-Commencing Master instance shutdown with mode='fast'
20170816:12:54:12:097793 gpstop:sdw1:gpadmin-[INFO]:-Master host=sdw1
20170816:12:54:12:097793 gpstop:sdw1:gpadmin-[INFO]:-Detected 0 connections to database
20170816:12:54:12:097793 gpstop:sdw1:gpadmin-[INFO]:-Using standard WAIT mode of 120 seconds
20170816:12:54:12:097793 gpstop:sdw1:gpadmin-[INFO]:-Commencing Master instance shutdown with mode=fast
20170816:12:54:12:097793 gpstop:sdw1:gpadmin-[INFO]:-Master segment instance directory=/hgdata/master/hgdwseg-1
20170816:12:54:13:097793 gpstop:sdw1:gpadmin-[INFO]:-Attempting forceful termination of any leftover master process
20170816:12:54:13:097793 gpstop:sdw1:gpadmin-[INFO]:-Terminating processes for segment /hgdata/master/hgdwseg-1
20170816:12:54:13:097793 gpstop:sdw1:gpadmin-[INFO]:-No standby master host configured
20170816:12:54:13:097793 gpstop:sdw1:gpadmin-[INFO]:-Commencing parallel segment instance shutdown, please wait...
20170816:12:54:13:097793 gpstop:sdw1:gpadmin-[INFO]:-0.00% of jobs completed
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-100.00% of jobs completed
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:- Segments stopped successfully = 12
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:- Segments with errors during stop = 0
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-Successfully shutdown 12 of 12 segment instances
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-Database successfully shutdown with no errors reported
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-Cleaning up leftover gpmmon process
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-No leftover gpmmon process found
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-Cleaning up leftover gpsmon processes
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-No leftover gpsmon processes on some hosts. not attempting forceful termination on these hosts
20170816:12:54:23:097793 gpstop:sdw1:gpadmin-[INFO]:-Cleaning up leftover shared memory
4.以管理模式启动数据库
[gpadmin@sdw1 ~]$ gpstart -m
20170816:12:54:40:098061 gpstart:sdw1:gpadmin-[INFO]:-Starting gpstart with args: -m
20170816:12:54:40:098061 gpstart:sdw1:gpadmin-[INFO]:-Gathering information and validating the environment...
20170816:12:54:40:098061 gpstart:sdw1:gpadmin-[INFO]:-Greenplum Binary Version: 'postgres (Greenplum Database) 4.3.99.00 build Deepgreen DB'
20170816:12:54:40:098061 gpstart:sdw1:gpadmin-[INFO]:-Greenplum Catalog Version: '201310150'
20170816:12:54:40:098061 gpstart:sdw1:gpadmin-[INFO]:-Master-only start requested in configuration without a standby master.
Continue with master-only startup Yy|Nn (default=N):
> y
20170816:12:54:41:098061 gpstart:sdw1:gpadmin-[INFO]:-Starting Master instance in admin mode
20170816:12:54:42:098061 gpstart:sdw1:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
20170816:12:54:42:098061 gpstart:sdw1:gpadmin-[INFO]:-Obtaining Segment details from master...
20170816:12:54:42:098061 gpstart:sdw1:gpadmin-[INFO]:-Setting new master era
20170816:12:54:42:098061 gpstart:sdw1:gpadmin-[INFO]:-Master Started...
5.登陆管理数据库
[gpadmin@sdw1 ~]$ PGOPTIONS="-c gp_session_role=utility" psql -d postgres
psql (8.2.15)
Type "help" for help.
6.删除segment
postgres=# select * from gp_segment_configuration;
dbid | content | role | preferred_role | mode | status | port | hostname | address | replication_port | san_mounts
------+---------+------+----------------+------+--------+-------+----------+---------+------------------+------------
1 | -1 | p | p | s | u | 5432 | sdw1 | sdw1 | |
2 | 0 | p | p | s | u | 25432 | sdw1 | sdw1 | |
3 | 1 | p | p | s | u | 25433 | sdw1 | sdw1 | |
4 | 2 | p | p | s | u | 25434 | sdw1 | sdw1 | |
5 | 3 | p | p | s | u | 25435 | sdw1 | sdw1 | |
6 | 4 | p | p | s | u | 25436 | sdw1 | sdw1 | |
7 | 5 | p | p | s | u | 25437 | sdw1 | sdw1 | |
8 | 6 | p | p | s | u | 25438 | sdw1 | sdw1 | |
9 | 7 | p | p | s | u | 25439 | sdw1 | sdw1 | |
10 | 8 | p | p | s | u | 25440 | sdw1 | sdw1 | |
11 | 9 | p | p | s | u | 25441 | sdw1 | sdw1 | |
12 | 10 | p | p | s | u | 25442 | sdw1 | sdw1 | |
13 | 11 | p | p | s | u | 25443 | sdw1 | sdw1 | |
(13 rows)
postgres=# set allow_system_table_mods='dml';
SET
postgres=# delete from gp_segment_configuration where dbid=13;
DELETE 1
postgres=# select * from gp_segment_configuration;
dbid | content | role | preferred_role | mode | status | port | hostname | address | replication_port | san_mounts
------+---------+------+----------------+------+--------+-------+----------+---------+------------------+------------
1 | -1 | p | p | s | u | 5432 | sdw1 | sdw1 | |
2 | 0 | p | p | s | u | 25432 | sdw1 | sdw1 | |
3 | 1 | p | p | s | u | 25433 | sdw1 | sdw1 | |
4 | 2 | p | p | s | u | 25434 | sdw1 | sdw1 | |
5 | 3 | p | p | s | u | 25435 | sdw1 | sdw1 | |
6 | 4 | p | p | s | u | 25436 | sdw1 | sdw1 | |
7 | 5 | p | p | s | u | 25437 | sdw1 | sdw1 | |
8 | 6 | p | p | s | u | 25438 | sdw1 | sdw1 | |
9 | 7 | p | p | s | u | 25439 | sdw1 | sdw1 | |
10 | 8 | p | p | s | u | 25440 | sdw1 | sdw1 | |
11 | 9 | p | p | s | u | 25441 | sdw1 | sdw1 | |
12 | 10 | p | p | s | u | 25442 | sdw1 | sdw1 | |
(12 rows)
7.删除filespace
postgres=# select * from pg_filespace_entry;
fsefsoid | fsedbid | fselocation
----------+---------+---------------------------
3052 | 1 | /hgdata/master/hgdwseg-1
3052 | 2 | /hgdata/primary/hgdwseg0
3052 | 3 | /hgdata/primary/hgdwseg1
3052 | 4 | /hgdata/primary/hgdwseg2
3052 | 5 | /hgdata/primary/hgdwseg3
3052 | 6 | /hgdata/primary/hgdwseg4
3052 | 7 | /hgdata/primary/hgdwseg5
3052 | 8 | /hgdata/primary/hgdwseg6
3052 | 9 | /hgdata/primary/hgdwseg7
3052 | 10 | /hgdata/primary/hgdwseg8
3052 | 11 | /hgdata/primary/hgdwseg9
3052 | 12 | /hgdata/primary/hgdwseg10
3052 | 13 | /hgdata/primary/hgdwseg11
(13 rows)
postgres=# delete from pg_filespace_entry where fsedbid=13;
DELETE 1
postgres=# select * from pg_filespace_entry;
fsefsoid | fsedbid | fselocation
----------+---------+---------------------------
3052 | 1 | /hgdata/master/hgdwseg-1
3052 | 2 | /hgdata/primary/hgdwseg0
3052 | 3 | /hgdata/primary/hgdwseg1
3052 | 4 | /hgdata/primary/hgdwseg2
3052 | 5 | /hgdata/primary/hgdwseg3
3052 | 6 | /hgdata/primary/hgdwseg4
3052 | 7 | /hgdata/primary/hgdwseg5
3052 | 8 | /hgdata/primary/hgdwseg6
3052 | 9 | /hgdata/primary/hgdwseg7
3052 | 10 | /hgdata/primary/hgdwseg8
3052 | 11 | /hgdata/primary/hgdwseg9
3052 | 12 | /hgdata/primary/hgdwseg10
(12 rows)
8.退出管理模式，正常启动数据库
[gpadmin@sdw1 ~]$ gpstop -m
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-Starting gpstop with args: -m
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-Gathering information and validating the environment...
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-Obtaining Segment details from master...
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-Greenplum Version: 'postgres (Greenplum Database) 4.3.99.00 build Deepgreen DB'
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-There are 0 connections to the database
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-Commencing Master instance shutdown with mode='smart'
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-Master host=sdw1
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-Commencing Master instance shutdown with mode=smart
20170816:12:56:52:098095 gpstop:sdw1:gpadmin-[INFO]:-Master segment instance directory=/hgdata/master/hgdwseg-1
20170816:12:56:53:098095 gpstop:sdw1:gpadmin-[INFO]:-Attempting forceful termination of any leftover master process
20170816:12:56:53:098095 gpstop:sdw1:gpadmin-[INFO]:-Terminating processes for segment /hgdata/master/hgdwseg-1
[gpadmin@sdw1 ~]$ gpstart
20170816:12:57:02:098112 gpstart:sdw1:gpadmin-[INFO]:-Starting gpstart with args:
20170816:12:57:02:098112 gpstart:sdw1:gpadmin-[INFO]:-Gathering information and validating the environment...
20170816:12:57:02:098112 gpstart:sdw1:gpadmin-[INFO]:-Greenplum Binary Version: 'postgres (Greenplum Database) 4.3.99.00 build Deepgreen DB'
20170816:12:57:02:098112 gpstart:sdw1:gpadmin-[INFO]:-Greenplum Catalog Version: '201310150'
20170816:12:57:02:098112 gpstart:sdw1:gpadmin-[INFO]:-Starting Master instance in admin mode
20170816:12:57:03:098112 gpstart:sdw1:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
20170816:12:57:03:098112 gpstart:sdw1:gpadmin-[INFO]:-Obtaining Segment details from master...
20170816:12:57:03:098112 gpstart:sdw1:gpadmin-[INFO]:-Setting new master era
20170816:12:57:03:098112 gpstart:sdw1:gpadmin-[INFO]:-Master Started...
20170816:12:57:03:098112 gpstart:sdw1:gpadmin-[INFO]:-Shutting down master
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:---------------------------
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:-Master instance parameters
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:---------------------------
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:-Database = template1
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:-Master Port = 5432
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:-Master directory = /hgdata/master/hgdwseg-1
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:-Timeout = 600 seconds
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:-Master standby = Off
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:---------------------------------------
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:-Segment instances that will be started
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:---------------------------------------
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- Host Datadir Port
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg0 25432
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg1 25433
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg2 25434
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg3 25435
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg4 25436
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg5 25437
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg6 25438
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg7 25439
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg8 25440
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg9 25441
20170816:12:57:05:098112 gpstart:sdw1:gpadmin-[INFO]:- sdw1 /hgdata/primary/hgdwseg10 25442
Continue with Greenplum instance startup Yy|Nn (default=N):
> y
20170816:12:57:07:098112 gpstart:sdw1:gpadmin-[INFO]:-Commencing parallel segment instance startup, please wait...
.......
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:-Process results...
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:- Successful segment starts = 11
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:- Failed segment starts = 0
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:- Skipped segment starts (segments are marked down in configuration) = 0
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:-
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:-Successfully started 11 of 11 segment instances
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:-----------------------------------------------------
20170816:12:57:14:098112 gpstart:sdw1:gpadmin-[INFO]:-Starting Master instance sdw1 directory /hgdata/master/hgdwseg-1
20170816:12:57:15:098112 gpstart:sdw1:gpadmin-[INFO]:-Command pg_ctl reports Master sdw1 instance active
20170816:12:57:15:098112 gpstart:sdw1:gpadmin-[INFO]:-No standby master configured. skipping...
20170816:12:57:15:098112 gpstart:sdw1:gpadmin-[INFO]:-Database successfully started
9.将删除节点的备份文件使用psql恢复到当前数据库
psql -d postgres -f xxxx.sql #这里不赘述恢复过程
备注：
1）本文使用的是只恢复删除节点的数据。
2）本文的过程，逆向执行，可以将删除的节点重新添加回来，但是数据恢复起来比较耗时，与重新建库恢复差不多。
转载自：https://www.sypopo.com/post/M95Rm39Or7/