当为segment节点配置了mirror时，如果segment节点挂了，FTS进程会监测到，并且自动将mirror节点切换为primary角色。
在数据库中可以查看到状态。下面表示优先角色为primary的节点已经down了。
postgres=# select * from gp_segment_configuration order by role,dbid;
dbid | content | role | preferred_role | mode | status | port | hostname | address | replication_port | san_mounts
------+---------+------+----------------+------+--------+-------+-------------------+-------------------+------------------+------------
2 | 0 | m | p | s | d | 40000 | digoal.sqa.zmf | digoal.sqa.zmf | 43000 |
3 | 1 | m | p | s | d | 40001 | digoal.sqa.zmf | digoal.sqa.zmf | 43001 |
4 | 2 | m | p | s | d | 40002 | digoal.sqa.zmf | digoal.sqa.zmf | 43002 |
5 | 3 | m | p | s | d | 40003 | digoal.sqa.zmf | digoal.sqa.zmf | 43003 |
6 | 4 | m | p | s | d | 40004 | digoal.sqa.zmf | digoal.sqa.zmf | 43004 |
7 | 5 | m | p | s | d | 40005 | digoal.sqa.zmf | digoal.sqa.zmf | 43005 |
8 | 6 | m | p | s | d | 40006 | digoal.sqa.zmf | digoal.sqa.zmf | 43006 |
9 | 7 | m | p | s | d | 40007 | digoal.sqa.zmf | digoal.sqa.zmf | 43007 |
10 | 8 | m | p | s | d | 40008 | digoal.sqa.zmf | digoal.sqa.zmf | 43008 |
11 | 9 | m | p | s | d | 40009 | digoal.sqa.zmf | digoal.sqa.zmf | 43009 |
12 | 10 | m | p | s | d | 40010 | digoal.sqa.zmf | digoal.sqa.zmf | 43010 |
13 | 11 | m | p | s | d | 40011 | digoal.sqa.zmf | digoal.sqa.zmf | 43011 |
14 | 12 | m | p | s | d | 40012 | digoal.sqa.zmf | digoal.sqa.zmf | 43012 |
15 | 13 | m | p | s | d | 40013 | digoal.sqa.zmf | digoal.sqa.zmf | 43013 |
16 | 14 | m | p | s | d | 40014 | digoal.sqa.zmf | digoal.sqa.zmf | 43014 |
17 | 15 | m | p | s | d | 40015 | digoal.sqa.zmf | digoal.sqa.zmf | 43015 |
18 | 16 | m | p | s | d | 40016 | digoal.sqa.zmf | digoal.sqa.zmf | 43016 |
19 | 17 | m | p | s | d | 40017 | digoal.sqa.zmf | digoal.sqa.zmf | 43017 |
20 | 18 | m | p | s | d | 40018 | digoal.sqa.zmf | digoal.sqa.zmf | 43018 |
21 | 19 | m | p | s | d | 40019 | digoal.sqa.zmf | digoal.sqa.zmf | 43019 |
22 | 20 | m | p | s | d | 40020 | digoal.sqa.zmf | digoal.sqa.zmf | 43020 |
23 | 21 | m | p | s | d | 40021 | digoal.sqa.zmf | digoal.sqa.zmf | 43021 |
24 | 22 | m | p | s | d | 50011 | digoal.sqa.zmf | digoal.sqa.zmf | 43022 |
1 | -1 | p | p | s | u | 1921 | digoal.sqa.zmf | digoal.sqa.zmf | |
25 | 0 | p | m | c | u | 41000 | digoal.sqa.zmf | digoal.sqa.zmf | 42000 |
26 | 1 | p | m | c | u | 41001 | digoal.sqa.zmf | digoal.sqa.zmf | 42001 |
27 | 2 | p | m | c | u | 41002 | digoal.sqa.zmf | digoal.sqa.zmf | 42002 |
28 | 3 | p | m | c | u | 41003 | digoal.sqa.zmf | digoal.sqa.zmf | 42003 |
29 | 4 | p | m | c | u | 41004 | digoal.sqa.zmf | digoal.sqa.zmf | 42004 |
30 | 5 | p | m | c | u | 41005 | digoal.sqa.zmf | digoal.sqa.zmf | 42005 |
31 | 6 | p | m | c | u | 41006 | digoal.sqa.zmf | digoal.sqa.zmf | 42006 |
32 | 7 | p | m | c | u | 41007 | digoal.sqa.zmf | digoal.sqa.zmf | 42007 |
33 | 8 | p | m | c | u | 41008 | digoal.sqa.zmf | digoal.sqa.zmf | 42008 |
34 | 9 | p | m | c | u | 41009 | digoal.sqa.zmf | digoal.sqa.zmf | 42009 |
35 | 10 | p | m | c | u | 41010 | digoal.sqa.zmf | digoal.sqa.zmf | 42010 |
36 | 11 | p | m | c | u | 41011 | digoal.sqa.zmf | digoal.sqa.zmf | 42011 |
37 | 12 | p | m | c | u | 41012 | digoal.sqa.zmf | digoal.sqa.zmf | 42012 |
38 | 13 | p | m | c | u | 41013 | digoal.sqa.zmf | digoal.sqa.zmf | 42013 |
39 | 14 | p | m | c | u | 41014 | digoal.sqa.zmf | digoal.sqa.zmf | 42014 |
40 | 15 | p | m | c | u | 41015 | digoal.sqa.zmf | digoal.sqa.zmf | 42015 |
41 | 16 | p | m | c | u | 41016 | digoal.sqa.zmf | digoal.sqa.zmf | 42016 |
42 | 17 | p | m | c | u | 41017 | digoal.sqa.zmf | digoal.sqa.zmf | 42017 |
43 | 18 | p | m | c | u | 41018 | digoal.sqa.zmf | digoal.sqa.zmf | 42018 |
44 | 19 | p | m | c | u | 41019 | digoal.sqa.zmf | digoal.sqa.zmf | 42019 |
45 | 20 | p | m | c | u | 41020 | digoal.sqa.zmf | digoal.sqa.zmf | 42020 |
46 | 21 | p | m | c | u | 41021 | digoal.sqa.zmf | digoal.sqa.zmf | 42021 |
47 | 22 | p | m | c | u | 41022 | digoal.sqa.zmf | digoal.sqa.zmf | 42022 |
(47 rows)
mode和status字段的含义。
mode = s, c, r 分别表示synced, change logging, resyning。
已同步，说明节点现在是mirror状态，并且已经完成同步。
跟踪块变更，说明节点现在是primary状态，正在产生块变更日志。
同步中，说明节点现在是mirror状态，并且还没有同步完成。
status = u, d 分布表示up 和 down。
preferred_role 表示角色优先为什么角色。
role表示当前为什么角色。
在这种状态下停库的话，会告诉你有几个节点已经DOWN了。
$gpstop -M fast -a
20151218:16:29:59:006401 gpstop:digoal:digoal-[INFO]:-----------------------------------------------------
20151218:16:29:59:006401 gpstop:digoal:digoal-[INFO]:- Segments stopped successfully = 46
20151218:16:29:59:006401 gpstop:digoal:digoal-[INFO]:- Segments with errors during stop = 0
20151218:16:29:59:006401 gpstop:digoal:digoal-[INFO]:-
20151218:16:29:59:006401 gpstop:digoal:digoal-[WARNING]:-Segments that are currently marked down in configuration = 23 <<<<<<<<
20151218:16:29:59:006401 gpstop:digoal:digoal-[INFO]:- (stop was still attempted on these segments)
启动数据库会忽略这些DOWN的节点
$gpstart
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg0 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg1 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg2 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg3 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg4 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg5 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg6 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg7 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg8 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg9 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg10 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg11 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg12 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg13 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg14 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg15 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg16 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg17 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg18 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg19 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg20 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg21 <<<<<
20151218:16:32:53:097351 gpstart:digoal:digoal-[WARNING]:-Skipping startup of segment marked down in configuration: on digoal.sqa.zmf directory /disk1/digoal/gpdata/gpseg22 <<<<<
使用gpstate -m可以看到，当前mirror处于primary角色。
$gpstate -m
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:-Starting gpstate with args: -m
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:-local Greenplum Version: 'postgres (Greenplum Database) 4.3.6.1 build 2'
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:-master Greenplum Version: 'PostgreSQL 8.2.15 (Greenplum Database 4.3.6.1 build 2) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Oct 1 2015 15:14:22'
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:-Obtaining Segment details from master...
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:--------------------------------------------------------------
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:--Current GPDB mirror list and status
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:--Type = Group
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:--------------------------------------------------------------
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- Mirror Datadir Port Status Data Status
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg0 41000 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg1 41001 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg2 41002 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg3 41003 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg4 41004 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg5 41005 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg6 41006 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg7 41007 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg8 41008 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg9 41009 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg10 41010 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg11 41011 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg12 41012 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg13 41013 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg14 41014 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg15 41015 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg16 41016 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg17 41017 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg18 41018 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg19 41019 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg20 41020 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg21 41021 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg22 41022 Acting as Primary Change Tracking
20151218:16:52:55:036249 gpstate:digoal:digoal-[INFO]:--------------------------------------------------------------
20151218:16:52:55:036249 gpstate:digoal:digoal-[WARNING]:-23 segment(s) configured as mirror(s) are acting as primaries
20151218:16:52:55:036249 gpstate:digoal:digoal-[WARNING]:-23 mirror segment(s) acting as primaries are in change tracking
进入主题，到底怎么恢复failed segment呢？
1. 首先要产生一个恢复配置文件
$gprecoverseg -o ./recov
20151218:17:02:03:034183 gprecoverseg:digoal:digoal-[INFO]:-Starting gprecoverseg with args: -o ./recov
20151218:17:02:03:034183 gprecoverseg:digoal:digoal-[INFO]:-local Greenplum Version: 'postgres (Greenplum Database) 4.3.6.1 build 2'
20151218:17:02:03:034183 gprecoverseg:digoal:digoal-[INFO]:-master Greenplum Version: 'PostgreSQL 8.2.15 (Greenplum Database 4.3.6.1 build 2) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Oct 1 2015 15:14:22'
20151218:17:02:03:034183 gprecoverseg:digoal:digoal-[INFO]:-Checking if segments are ready
20151218:17:02:03:034183 gprecoverseg:digoal:digoal-[INFO]:-Obtaining Segment details from master...
20151218:17:02:04:034183 gprecoverseg:digoal:digoal-[INFO]:-Obtaining Segment details from master...
20151218:17:02:07:034183 gprecoverseg:digoal:digoal-[INFO]:-Configuration file output to ./recov successfully.
配置文件内容
$cat ./recov
filespaceOrder=
digoal.sqa.zmf:40000:/disk1/digoal/gpdata/gpseg0
digoal.sqa.zmf:40001:/disk1/digoal/gpdata/gpseg1
digoal.sqa.zmf:40002:/disk1/digoal/gpdata/gpseg2
digoal.sqa.zmf:40003:/disk1/digoal/gpdata/gpseg3
digoal.sqa.zmf:40004:/disk1/digoal/gpdata/gpseg4
digoal.sqa.zmf:40005:/disk1/digoal/gpdata/gpseg5
digoal.sqa.zmf:40006:/disk1/digoal/gpdata/gpseg6
digoal.sqa.zmf:40007:/disk1/digoal/gpdata/gpseg7
digoal.sqa.zmf:40008:/disk1/digoal/gpdata/gpseg8
digoal.sqa.zmf:40009:/disk1/digoal/gpdata/gpseg9
digoal.sqa.zmf:40010:/disk1/digoal/gpdata/gpseg10
digoal.sqa.zmf:40011:/disk1/digoal/gpdata/gpseg11
digoal.sqa.zmf:40012:/disk1/digoal/gpdata/gpseg12
digoal.sqa.zmf:40013:/disk1/digoal/gpdata/gpseg13
digoal.sqa.zmf:40014:/disk1/digoal/gpdata/gpseg14
digoal.sqa.zmf:40015:/disk1/digoal/gpdata/gpseg15
digoal.sqa.zmf:40016:/disk1/digoal/gpdata/gpseg16
digoal.sqa.zmf:40017:/disk1/digoal/gpdata/gpseg17
digoal.sqa.zmf:40018:/disk1/digoal/gpdata/gpseg18
digoal.sqa.zmf:40019:/disk1/digoal/gpdata/gpseg19
digoal.sqa.zmf:40020:/disk1/digoal/gpdata/gpseg20
digoal.sqa.zmf:40021:/disk1/digoal/gpdata/gpseg21
digoal.sqa.zmf:50011:/disk1/digoal/gpdata/gpseg22
2. 然后使用这个配置文件进行恢复
$gprecoverseg -i ./recov
。。。
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Synchronization mode = Incremental
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Failed instance host = digoal.sqa.zmf
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Failed instance address = digoal.sqa.zmf
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Failed instance directory = /disk1/digoal/gpdata/gpseg22
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Failed instance port = 50011
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Failed instance replication port = 43022
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Recovery Source instance host = digoal.sqa.zmf
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Recovery Source instance address = digoal.sqa.zmf
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Recovery Source instance directory = /disk1/digoal/gpdata_mirror/gpseg22
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Recovery Source instance port = 41022
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Recovery Source instance replication port = 42022
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:- Recovery Target = in-place
20151218:17:02:28:050151 gprecoverseg:digoal:digoal-[INFO]:----------------------------------------------------------
Continue with segment recovery procedure Yy|Nn (default=N):
> y
恢复过程中，查看mirror状态。
$gpstate -m
...
20151218:17:03:00:072483 gpstate:digoal:digoal-[INFO]:- Mirror Datadir Port Status Data Status
20151218:17:03:00:072483 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg0 41000 Acting as Primary Resynchronizing
20151218:17:03:00:072483 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg1 41001 Acting as Primary Resynchronizing
20151218:17:03:00:072483 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg2 41002 Acting as Primary Resynchronizing
...
同步完成的状态如下 （Synchronized）
20151218:17:08:38:106220 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg1 41001 Acting as Primary Synchronized
20151218:17:08:38:106220 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg2 41002 Acting as Primary Synchronized
20151218:17:08:38:106220 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg3 41003 Acting as Primary Synchronized
3. 到上一步，数据库的主备就恢复了，但是还有一步，是可选的。
你要不要把primary , mirror角色对调一下，因为现在mirror和primary和优先角色是相反的。
如果要对调，使用以下命令，会停库来处理。
$gprecoverseg -r
20151218:17:12:16:120198 gprecoverseg:digoal:digoal-[INFO]:-Starting gprecoverseg with args: -r
20151218:17:12:16:120198 gprecoverseg:digoal:digoal-[INFO]:-local Greenplum Version: 'postgres (Greenplum Database) 4.3.6.1 build 2'
20151218:17:12:16:120198 gprecoverseg:digoal:digoal-[INFO]:-master Greenplum Version: 'PostgreSQL 8.2.15 (Greenplum Database 4.3.6.1 build 2) on x86_64-unknown-linux-gnu, compiled by GCC gcc (GCC) 4.4.2 compiled on Oct 1 2015 15:14:22'
20151218:17:12:16:120198 gprecoverseg:digoal:digoal-[INFO]:-Checking if segments are ready
20151218:17:12:16:120198 gprecoverseg:digoal:digoal-[INFO]:-Obtaining Segment details from master...
20151218:17:12:17:120198 gprecoverseg:digoal:digoal-[INFO]:-Obtaining Segment details from master...
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:-Greenplum instance recovery parameters
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:----------------------------------------------------------
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:-Recovery type = Rebalance
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:----------------------------------------------------------
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:-Unbalanced segment 1 of 46
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:----------------------------------------------------------
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:- Unbalanced instance host = digoal.sqa.zmf
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:- Unbalanced instance address = digoal.sqa.zmf
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:- Unbalanced instance directory = /disk1/digoal/gpdata_mirror/gpseg0
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:- Unbalanced instance port = 41000
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:- Unbalanced instance replication port = 42000
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:- Balanced role = Mirror
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:- Current role = Primary
......
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[INFO]:----------------------------------------------------------
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[WARNING]:-This operation will cancel queries that are currently executing.
20151218:17:12:18:120198 gprecoverseg:digoal:digoal-[WARNING]:-Connections to the database however will not be interrupted.
Continue with segment rebalance procedure Yy|Nn (default=N):
> y
...
20151218:17:13:03:120198 gprecoverseg:digoal:digoal-[INFO]:-******************************************************************
20151218:17:13:03:120198 gprecoverseg:digoal:digoal-[INFO]:-The rebalance operation has completed successfully.
20151218:17:13:03:120198 gprecoverseg:digoal:digoal-[INFO]:-There is a resynchronization running in the background to bring all
20151218:17:13:03:120198 gprecoverseg:digoal:digoal-[INFO]:-segments in sync.
20151218:17:13:03:120198 gprecoverseg:digoal:digoal-[INFO]:-
20151218:17:13:03:120198 gprecoverseg:digoal:digoal-[INFO]:-Use gpstate -e to check the resynchronization progress.
20151218:17:13:03:120198 gprecoverseg:digoal:digoal-[INFO]:-******************************************************************
查看同步状态
$gpstate -m
。。。
20151218:17:16:09:093083 gpstate:digoal:digoal-[INFO]:--------------------------------------------------------------
20151218:17:16:09:093083 gpstate:digoal:digoal-[INFO]:- Mirror Datadir Port Status Data Status
20151218:17:16:09:093083 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg0 41000 Passive Synchronized
20151218:17:16:09:093083 gpstate:digoal:digoal-[INFO]:- digoal.sqa.zmf /disk1/digoal/gpdata_mirror/gpseg1 41001 Passive Synchronized
。。。
查看segment配置，已经是synced和up的状态了。
postgres=# select * from gp_segment_configuration order by role,dbid;
dbid | content | role | preferred_role | mode | status | port | hostname | address | replication_port | san_mounts
------+---------+------+----------------+------+--------+-------+-------------------+-------------------+------------------+------------
25 | 0 | m | m | s | u | 41000 | digoal.sqa.zmf | digoal.sqa.zmf | 42000 |
26 | 1 | m | m | s | u | 41001 | digoal.sqa.zmf | digoal.sqa.zmf | 42001 |
27 | 2 | m | m | s | u | 41002 | digoal.sqa.zmf | digoal.sqa.zmf | 42002 |
28 | 3 | m | m | s | u | 41003 | digoal.sqa.zmf | digoal.sqa.zmf | 42003 |
29 | 4 | m | m | s | u | 41004 | digoal.sqa.zmf | digoal.sqa.zmf | 42004 |
30 | 5 | m | m | s | u | 41005 | digoal.sqa.zmf | digoal.sqa.zmf | 42005 |
31 | 6 | m | m | s | u | 41006 | digoal.sqa.zmf | digoal.sqa.zmf | 42006 |
32 | 7 | m | m | s | u | 41007 | digoal.sqa.zmf | digoal.sqa.zmf | 42007 |
33 | 8 | m | m | s | u | 41008 | digoal.sqa.zmf | digoal.sqa.zmf | 42008 |
34 | 9 | m | m | s | u | 41009 | digoal.sqa.zmf | digoal.sqa.zmf | 42009 |
35 | 10 | m | m | s | u | 41010 | digoal.sqa.zmf | digoal.sqa.zmf | 42010 |
36 | 11 | m | m | s | u | 41011 | digoal.sqa.zmf | digoal.sqa.zmf | 42011 |
37 | 12 | m | m | s | u | 41012 | digoal.sqa.zmf | digoal.sqa.zmf | 42012 |
38 | 13 | m | m | s | u | 41013 | digoal.sqa.zmf | digoal.sqa.zmf | 42013 |
39 | 14 | m | m | s | u | 41014 | digoal.sqa.zmf | digoal.sqa.zmf | 42014 |
40 | 15 | m | m | s | u | 41015 | digoal.sqa.zmf | digoal.sqa.zmf | 42015 |
41 | 16 | m | m | s | u | 41016 | digoal.sqa.zmf | digoal.sqa.zmf | 42016 |
42 | 17 | m | m | s | u | 41017 | digoal.sqa.zmf | digoal.sqa.zmf | 42017 |
43 | 18 | m | m | s | u | 41018 | digoal.sqa.zmf | digoal.sqa.zmf | 42018 |
44 | 19 | m | m | s | u | 41019 | digoal.sqa.zmf | digoal.sqa.zmf | 42019 |
45 | 20 | m | m | s | u | 41020 | digoal.sqa.zmf | digoal.sqa.zmf | 42020 |
46 | 21 | m | m | s | u | 41021 | digoal.sqa.zmf | digoal.sqa.zmf | 42021 |
47 | 22 | m | m | s | u | 41022 | digoal.sqa.zmf | digoal.sqa.zmf | 42022 |
1 | -1 | p | p | s | u | 1921 | digoal.sqa.zmf | digoal.sqa.zmf | |
2 | 0 | p | p | s | u | 40000 | digoal.sqa.zmf | digoal.sqa.zmf | 43000 |
3 | 1 | p | p | s | u | 40001 | digoal.sqa.zmf | digoal.sqa.zmf | 43001 |
4 | 2 | p | p | s | u | 40002 | digoal.sqa.zmf | digoal.sqa.zmf | 43002 |
5 | 3 | p | p | s | u | 40003 | digoal.sqa.zmf | digoal.sqa.zmf | 43003 |
6 | 4 | p | p | s | u | 40004 | digoal.sqa.zmf | digoal.sqa.zmf | 43004 |
7 | 5 | p | p | s | u | 40005 | digoal.sqa.zmf | digoal.sqa.zmf | 43005 |
8 | 6 | p | p | s | u | 40006 | digoal.sqa.zmf | digoal.sqa.zmf | 43006 |
9 | 7 | p | p | s | u | 40007 | digoal.sqa.zmf | digoal.sqa.zmf | 43007 |
10 | 8 | p | p | s | u | 40008 | digoal.sqa.zmf | digoal.sqa.zmf | 43008 |
11 | 9 | p | p | s | u | 40009 | digoal.sqa.zmf | digoal.sqa.zmf | 43009 |
12 | 10 | p | p | s | u | 40010 | digoal.sqa.zmf | digoal.sqa.zmf | 43010 |
13 | 11 | p | p | s | u | 40011 | digoal.sqa.zmf | digoal.sqa.zmf | 43011 |
14 | 12 | p | p | s | u | 40012 | digoal.sqa.zmf | digoal.sqa.zmf | 43012 |
15 | 13 | p | p | s | u | 40013 | digoal.sqa.zmf | digoal.sqa.zmf | 43013 |
16 | 14 | p | p | s | u | 40014 | digoal.sqa.zmf | digoal.sqa.zmf | 43014 |
17 | 15 | p | p | s | u | 40015 | digoal.sqa.zmf | digoal.sqa.zmf | 43015 |
18 | 16 | p | p | s | u | 40016 | digoal.sqa.zmf | digoal.sqa.zmf | 43016 |
19 | 17 | p | p | s | u | 40017 | digoal.sqa.zmf | digoal.sqa.zmf | 43017 |
20 | 18 | p | p | s | u | 40018 | digoal.sqa.zmf | digoal.sqa.zmf | 43018 |
21 | 19 | p | p | s | u | 40019 | digoal.sqa.zmf | digoal.sqa.zmf | 43019 |
22 | 20 | p | p | s | u | 40020 | digoal.sqa.zmf | digoal.sqa.zmf | 43020 |
23 | 21 | p | p | s | u | 40021 | digoal.sqa.zmf | digoal.sqa.zmf | 43021 |
24 | 22 | p | p | s | u | 50011 | digoal.sqa.zmf | digoal.sqa.zmf | 43022 |
(47 rows)
[参考]
gprecoverseg -h
Greenplum failed segment的恢复方法
2015-12-18
7117
版权
版权声明：
本文内容由阿里云实名注册用户自发贡献，版权归原作者所有，阿里云开发者社区不拥有其著作权，亦不承担相应法律责任。具体规则请查看《
阿里云开发者社区用户服务协议》和
《阿里云开发者社区知识产权保护指引》。如果您发现本社区中有涉嫌抄袭的内容，填写
侵权投诉表单进行举报，一经查实，本社区将立刻删除涉嫌侵权内容。
简介：
目录
相关文章
|
6月前
|
Oracle
关系型数据库
MySQL
实时计算 Flink版操作报错之使用oracle-cdc的，遇到错误：ORA-01292: no log file has been specified for the current LogMiner session，该如何处理
在使用实时计算Flink版过程中，可能会遇到各种错误，了解这些错误的原因及解决方法对于高效排错至关重要。针对具体问题，查看Flink的日志是关键，它们通常会提供更详细的错误信息和堆栈跟踪，有助于定位问题。此外，Flink社区文档和官方论坛也是寻求帮助的好去处。以下是一些常见的操作报错及其可能的原因与解决策略。
244
0
0
|
SQL
监控
关系型数据库
|
存储
关系型数据库
MySQL
mysql 数据库无法启动（Ignoring the redo log due to missing MLOG_CHECKPOINT between the checkpoint .... and)
数据库机器的CPU和主板都换了，重新开机，发现mysql数据库无法启动！
343
0
0
|
关系型数据库
数据库
PostgreSQL
PG异常无法启动的问题：could not read file "pg_logical/replorigin_checkpoint": Success
问题描述
新安装不久的PostgreSQL数据库，断电后重启，查看日志如下
2019-01-08 08:44:19.989 UTC [7493] LOG: database system was interrupted; last known up at 2018-12-24 10:56:28 UTC
2019-01-08 08:44:19.
3384
0
1
|
SQL
关系型数据库
Java
|
容灾
关系型数据库
数据库
|
SQL
Oracle
关系型数据库
xDB Replication Server - PostgreSQL, Oracle, SQL Server, PPAS 全量、增量(redo log based, or trigger based)同步（支持single-master, mult-master同步, 支持DDL）
xDB Replication Server - PostgreSQL, Oracle, SQL Server, PPAS 全量、增量(redo log based, or trigger based)同步（支持single-master, mult-master同步, 支持DDL）
852
0
0
|
SQL
Oracle
关系型数据库
PostgreSQL 物化视图(Oracle同步到PG，PG同步到PG) - by pgsnapshot (plperlu trigger) （支持类似Oracle的mvlog fast complete force刷新）
标签
PostgreSQL , 物化视图 , 增量刷新 , mvlog , Oracle 同步到 PG , PG 同步到 PG
背景
PostgreSQL自身的物化视图没有MVLOG，也就是说，刷新的时候是VIEW定义产生的记录与MV已刷新的记录进行比对，进行增量更新的过程。
3332
0
0
|
关系型数据库
PostgreSQL
|
关系型数据库
PostgreSQL
热门文章
最新文章
1
Object reference not set to an instance of an object.
2
Map根据键、值进行排序
3
锐捷网络&视易K米，携手助力2016腾讯全球合作伙伴万人大会
4
Java中的定时器
5
Python正则表达式指南下下半部
6
VirtualDesktop多媒体在何处解码?
7
文件移动脚本
8
xml在此生活
9
在不动用sp_configure的情况下，如何 =》去掉列的自增长，并保留原数据
10
聚焦技术实战！MDCC 2016 移动开发者大会盛大开幕
1
2024Mysql And Redis基础与进阶操作系列（6）作者——LJS[含MySQL 多表之一对一/多；多对多；多表联合查询等详解步骤及常见报错问题所对应的解决方法]
36
2
猫头虎分享：鸿蒙生态带给开发者的全新机遇！轻松实现按需加载与多端适配，开发效率翻倍
34
3
2024Mysql And Redis基础与进阶操作系列（5）作者——LJS[含MySQL DQL基本查询:select；简单、排序、分组、聚合、分组、分页等详解步骤及常见报错问题所对应的解决方法]
28
4
理解 ES6 中的 Promise
27
5
CSS 提高性能的方法
23
6
CSS 预编语言的区别
10
7
如何快速掌握 CSS 预编语言
10
8
对 CSS 预编语言的理解
12
9
利用缓存布局信息来减少回流和重绘的发生
12
10
减少回流和重绘的发生
12