Greenplum删除segment节点

1 备份数据库

[gpadmin@hadoop02 ~]$ gp_dump --gp-d=/home/gpadmin/backup tutorial
20160928:14:41:34|gp_dump-[INFO]:-Read params: <empty>
20160928:14:41:34|gp_dump-[INFO]:-Command line options analyzed.
20160928:14:41:34|gp_dump-[INFO]:-Connecting to master database on host localhost port 5432 database tutorial.
20160928:14:41:34|gp_dump-[INFO]:-Reading Greenplum Database configuration info from master database.
20160928:14:41:34|gp_dump-[INFO]:-Preparing to dump the following segments:
20160928:14:41:34|gp_dump-[INFO]:-Segment 1 (dbid 3)
20160928:14:41:34|gp_dump-[INFO]:-Segment 0 (dbid 2)
20160928:14:41:34|gp_dump-[INFO]:-Master (dbid 1)
20160928:14:41:34|gp_dump-[INFO]:-Starting a transaction on master database tutorial.
20160928:14:41:34|gp_dump-[INFO]:-Getting a lock on pg_class in database tutorial.
20160928:14:41:34|GetTimestampKey-[INFO]:-Timestamp key is generated as it is not provided by the user.
20160928:14:41:34|gp_dump-[INFO]:-About to spin off 3 threads with timestamp key 20160928144134
20160928:14:41:34|gp_dump-[INFO]:-Creating thread to backup dbid 3: host hadoop07 port 40000 database tutorial
20160928:14:41:34|gp_dump-[INFO]:-Creating thread to backup dbid 2: host hadoop04 port 40000 database tutorial
20160928:14:41:34|gp_dump-[INFO]:-Creating thread to backup dbid 1: host hadoop02 port 5432 database tutorial
20160928:14:41:34|gp_dump-[INFO]:-Waiting for remote gp_dump_agent processes to start transactions in serializable isolation level
20160928:14:41:34|gp_dump-[INFO]:-Listening for messages from server on dbid 1 connection
20160928:14:41:34|gp_dump-[INFO]:-Listening for messages from server on dbid 2 connection
20160928:14:41:34|gp_dump-[INFO]:-Listening for messages from server on dbid 3 connection
20160928:14:41:34|gp_dump-[INFO]:-Successfully launched Greenplum Database backup on dbid 2 server
20160928:14:41:34|gp_dump-[INFO]:-Successfully launched Greenplum Database backup on dbid 3 server
20160928:14:41:36|gp_dump-[INFO]:-Successfully launched Greenplum Database backup on dbid 1 server
20160928:14:41:36|gp_dump-[INFO]:-backup succeeded for dbid 1 on host hadoop02
20160928:14:41:36|gp_dump-[INFO]:-All remote gp_dump_agent processes have began transactions in serializable isolation level
20160928:14:41:36|gp_dump-[INFO]:-Waiting for remote gp_dump_agent processes to obtain local locks on dumpable objects
20160928:14:41:36|gp_dump-[INFO]:-All remote gp_dump_agent processes have obtains the necessary locks
20160928:14:41:36|gp_dump-[INFO]:-Committing transaction on the master database, thereby releasing locks.
20160928:14:41:36|gp_dump-[INFO]:-Waiting for all remote gp_dump_agent programs to finish.
20160928:14:42:00|gp_dump-[INFO]:-backup succeeded for dbid 3 on host hadoop07
20160928:14:42:00|gp_dump-[INFO]:-backup succeeded for dbid 2 on host hadoop04
20160928:14:42:00|gp_dump-[INFO]:-All remote gp_dump_agent programs are finished.
20160928:14:42:00|gp_dump-[INFO]:-Report results also written to /data/master/gpseg-1/gp_dump_20160928144134.rpt.
 
Greenplum Database Backup Report
Timestamp Key: 20160928144134
gp_dump Command Line: --gp-d=/home/gpadmin/backup tutorial
Pass through Command Line Options: None
Compression Program: None
Backup Type: Full
 
Individual Results
segment 1 (dbid 3) Host hadoop07 Port 40000 Database tutorial BackupFile /home/gpadmin/backup/gp_dump_0_3_20160928144134: Succeeded
segment 0 (dbid 2) Host hadoop04 Port 40000 Database tutorial BackupFile /home/gpadmin/backup/gp_dump_0_2_20160928144134: Succeeded
Master (dbid 1) Host hadoop02 Port 5432 Database tutorial BackupFile /home/gpadmin/backup/gp_dump_1_1_20160928144134: Succeeded
Master (dbid 1) Host hadoop02 Port 5432 Database tutorial BackupFile /home/gpadmin/backup/gp_dump_1_1_20160928144134_post_data: Succeeded
 
gp_dump utility finished successfully.
 

2 关闭数据库

[gpadmin@hadoop02 ~]$ gpstop -M fast -a
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Starting gpstop with args: -M fast -a
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Gathering information and validating the environment...
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Obtaining Segment details from master...
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Greenplum Version: 'postgres (Greenplum Database) 4.3.8.1 build 1'
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-There are 0 connections to the database
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Commencing Master instance shutdown with mode='fast'
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Master host=hadoop02
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Detected 0 connections to database
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Using standard WAIT mode of 120 seconds
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Commencing Master instance shutdown with mode=fast
20160928:14:43:39:002897 gpstop:hadoop02:gpadmin-[INFO]:-Master segment instance directory=/data/master/gpseg-1
20160928:14:43:40:002897 gpstop:hadoop02:gpadmin-[INFO]:-Attempting forceful termination of any leftover master process
20160928:14:43:40:002897 gpstop:hadoop02:gpadmin-[INFO]:-Terminating processes for segment /data/master/gpseg-1
20160928:14:43:40:002897 gpstop:hadoop02:gpadmin-[INFO]:-Stopping master standby host hadoop03 mode=fast
20160928:14:43:41:002897 gpstop:hadoop02:gpadmin-[INFO]:-Successfully shutdown standby process on hadoop03
20160928:14:43:41:002897 gpstop:hadoop02:gpadmin-[INFO]:-Commencing parallel segment instance shutdown, please wait...
20160928:14:43:41:002897 gpstop:hadoop02:gpadmin-[INFO]:-0.00% of jobs completed
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-100.00% of jobs completed
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-----------------------------------------------------
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-   Segments stopped successfully      = 2
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-   Segments with errors during stop   = 0
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-----------------------------------------------------
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-Successfully shutdown 2 of 2 segment instances
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-Database successfully shutdown with no errors reported
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-Cleaning up leftover gpmmon process
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-No leftover gpmmon process found
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-Cleaning up leftover gpsmon processes
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-No leftover gpsmon processes on some hosts. not attempting forceful termination on these hosts
20160928:14:43:51:002897 gpstop:hadoop02:gpadmin-[INFO]:-Cleaning up leftover shared memory

 

3 master模式启动数据库

[gpadmin@hadoop02 ~]$ gpstart -m
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[INFO]:-Starting gpstart with args: -m
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[INFO]:-Gathering information and validating the environment...
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[INFO]:-Greenplum Binary Version: 'postgres (Greenplum Database) 4.3.8.1 build 1'
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[INFO]:-Greenplum Catalog Version: '201310150'
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[WARNING]:-****************************************************************************
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[WARNING]:-Master-only start requested in a configuration with a standby master.
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[WARNING]:-This is advisable only under the direct supervision of Greenplum support.
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[WARNING]:-This mode of operation is not supported in a production environment and
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[WARNING]:-may lead to a split-brain condition and possible unrecoverable data loss.
20160928:14:45:01:003022 gpstart:hadoop02:gpadmin-[WARNING]:-****************************************************************************
 
Continue with master-only startup Yy|Nn (default=N):
> y
20160928:14:45:03:003022 gpstart:hadoop02:gpadmin-[INFO]:-Starting Master instance in admin mode
20160928:14:45:04:003022 gpstart:hadoop02:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
20160928:14:45:04:003022 gpstart:hadoop02:gpadmin-[INFO]:-Obtaining Segment details from master...
20160928:14:45:04:003022 gpstart:hadoop02:gpadmin-[INFO]:-Setting new master era
20160928:14:45:04:003022 gpstart:hadoop02:gpadmin-[INFO]:-Master Started...
 

4 进入管理模式修改元数据

[gpadmin@hadoop02 ~]$ PGOPTIONS="-c gp_session_role=utility" psql
psql (8.2.15)
Type "help" for help.
gpadmin=# 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 | hadoop02 | hadoop02 |                  |
    2 |       0 | p    | p              | s    | u      | 40000 | hadoop04 | hadoop04 |                  |
    3 |       1 | p    | p              | s    | u      | 40000 | hadoop07 | hadoop07 |                  |
    4 |      -1 | m    | m              | s    | u      |  5432 | hadoop03 | hadoop03 |                  |
(4 rows)
 
gpadmin=# set allow_system_table_mods='dml'; --获取修改系统表的权限
SET
 
gpadmin=# delete from gp_segment_configuration where dbid=3;                                                                         
DELETE 1
gpadmin=# 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 | hadoop02 | hadoop02 |                  |
    2 |       0 | p    | p              | s    | u      | 40000 | hadoop04 | hadoop04 |                  |
    4 |      -1 | m    | m              | s    | u      |  5432 | hadoop03 | hadoop03 |                  |
(3 rows)
 
gpadmin=# select * from pg_filespace_entry;
 fsefsoid | fsedbid |     fselocation      
----------+---------+----------------------
     3052 |       1 | /data/master/gpseg-1
     3052 |       2 | /data/primary/gpseg0
     3052 |       3 | /data/primary/gpseg1
     3052 |       4 | /data/master/gpseg-1
(4 rows)
 
gpadmin=# delete from pg_filespace_entry where fsedbid=3;
DELETE 1
gpadmin=# select * from pg_filespace_entry;
 fsefsoid | fsedbid |     fselocation      
----------+---------+----------------------
     3052 |       1 | /data/master/gpseg-1
     3052 |       2 | /data/primary/gpseg0
     3052 |       4 | /data/master/gpseg-1
(3 rows)

 

 

5 集群模式启动数据库

5.1 关闭master模式启动的gp

[gpadmin@hadoop02 ~]$ gpstop -m
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-Starting gpstop with args: -m
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-Gathering information and validating the environment...
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-Obtaining Segment details from master...
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-Greenplum Version: 'postgres (Greenplum Database) 4.3.8.1 build 1'
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-There are 0 connections to the database
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-Commencing Master instance shutdown with mode='smart'
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-Master host=hadoop02
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-Commencing Master instance shutdown with mode=smart
20160928:14:52:12:003181 gpstop:hadoop02:gpadmin-[INFO]:-Master segment instance directory=/data/master/gpseg-1
20160928:14:52:13:003181 gpstop:hadoop02:gpadmin-[INFO]:-Attempting forceful termination of any leftover master process
20160928:14:52:13:003181 gpstop:hadoop02:gpadmin-[INFO]:-Terminating processes for segment /data/master/gpseg-1

 

5.2 启动数据库

[gpadmin@hadoop02 ~]$ gpstart -a
20160928:14:52:53:003210 gpstart:hadoop02:gpadmin-[INFO]:-Starting gpstart with args: -a
20160928:14:52:53:003210 gpstart:hadoop02:gpadmin-[INFO]:-Gathering information and validating the environment...
20160928:14:52:53:003210 gpstart:hadoop02:gpadmin-[INFO]:-Greenplum Binary Version: 'postgres (Greenplum Database) 4.3.8.1 build 1'
20160928:14:52:53:003210 gpstart:hadoop02:gpadmin-[INFO]:-Greenplum Catalog Version: '201310150'
20160928:14:52:53:003210 gpstart:hadoop02:gpadmin-[INFO]:-Starting Master instance in admin mode
20160928:14:52:54:003210 gpstart:hadoop02:gpadmin-[INFO]:-Obtaining Greenplum Master catalog information
20160928:14:52:54:003210 gpstart:hadoop02:gpadmin-[INFO]:-Obtaining Segment details from master...
20160928:14:52:54:003210 gpstart:hadoop02:gpadmin-[INFO]:-Setting new master era
20160928:14:52:54:003210 gpstart:hadoop02:gpadmin-[INFO]:-Master Started...
20160928:14:52:54:003210 gpstart:hadoop02:gpadmin-[INFO]:-Shutting down master
20160928:14:52:55:003210 gpstart:hadoop02:gpadmin-[INFO]:-Commencing parallel segment instance startup, please wait...
..
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-Process results...
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-----------------------------------------------------
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-   Successful segment starts                                            = 1
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-   Failed segment starts                                                = 0
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-   Skipped segment starts (segments are marked down in configuration)   = 0
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-----------------------------------------------------
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-Successfully started 1 of 1 segment instances
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-----------------------------------------------------
20160928:14:52:57:003210 gpstart:hadoop02:gpadmin-[INFO]:-Starting Master instance hadoop02 directory /data/master/gpseg-1
20160928:14:52:58:003210 gpstart:hadoop02:gpadmin-[INFO]:-Command pg_ctl reports Master hadoop02 instance active
20160928:14:52:59:003210 gpstart:hadoop02:gpadmin-[INFO]:-Starting standby master
20160928:14:52:59:003210 gpstart:hadoop02:gpadmin-[INFO]:-Checking if standby master is running on host: hadoop03  in directory: /data/master/gpseg-1
20160928:14:52:59:003210 gpstart:hadoop02:gpadmin-[INFO]:-Database successfully started

 

6 验证节点是否正确删除

[gpadmin@hadoop02 ~]$ gpstate -s
20160928:14:54:01:003347 gpstate:hadoop02:gpadmin-[INFO]:-Starting gpstate with args: -s
20160928:14:54:01:003347 gpstate:hadoop02:gpadmin-[INFO]:-local Greenplum Version: 'postgres (Greenplum Database) 4.3.8.1 build 1'
20160928:14:54:01:003347 gpstate:hadoop02:gpadmin-[INFO]:-master Greenplum Version: 'PostgreSQL 8.2.15 (Greenplum Database 4.3.8.1 build 1) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Apr 20 2016 08:08:56'
20160928:14:54:01:003347 gpstate:hadoop02:gpadmin-[INFO]:-Obtaining Segment details from master...
20160928:14:54:01:003347 gpstate:hadoop02:gpadmin-[INFO]:-Gathering data from segments...
.
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-----------------------------------------------------
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:--Master Configuration & Status
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-----------------------------------------------------
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Master host                    = hadoop02
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Master postgres process ID     = 3266
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Master data directory          = /data/master/gpseg-1
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Master port                    = 5432
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Master current role            = dispatch
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Greenplum initsystem version   = 4.3.8.1 build 1
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Greenplum current version      = PostgreSQL 8.2.15 (Greenplum Database 4.3.8.1 build 1) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Apr 20 2016 08:08:56
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Postgres version               = 8.2.15
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Master standby                 = hadoop03
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Standby master state           = Standby host passive
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-----------------------------------------------------
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-Segment Instance Status Report
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-----------------------------------------------------
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Segment Info
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-      Hostname                          = hadoop04
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-      Address                           = hadoop04
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-      Datadir                           = /data/primary/gpseg0
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-      Port                              = 40000
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-   Status
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-      PID                               = 1489
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-      Configuration reports status as   = Up
20160928:14:54:02:003347 gpstate:hadoop02:gpadmin-[INFO]:-      Database status                   = Up

 

7 恢复数据

[gpadmin@hadoop07 backup]$ pwd

/home/gpadmin/backup

[gpadmin@hadoop07 backup]$ ls

gp_dump_0_3_20160928144134  gp_dump_status_0_3_20160928144134

 

恢复删除节点上的数据，只需将删除掉的节点上的数据重分布即可，gp_dump_0_3_20160928144134是保存在hadoop07 上的，先将其传到那hadoop02上，再执行以下命令。

 

[gpadmin@hadoop02 ~]$ psql tutorial -f gp_dump_0_3_20160928144134

SET

SET

SET

SET

SET

SET

SET

SET

 setval

--------

      1

(1 row)

 

至此节点删除完成