v953跨集群gcrcman备份恢复测试

一、编写目的

该测试方案针对v953版本gcrcman跨集群备份测试
二、环境说明
测试环境redhat7.7,集群环境为3节点，1分片，1备份。
三、测试方案
以下统称备份集群为源集群，恢复集群为目标集群
1、源集群和目标集群IP地址不同，管理节点和数据节点数和分片数一致，distribution id不同
2、源集群和目标集群ip地址相同，管理节点和数据节点数和分片数一致，distribution id不同
四、测试版本
GBase8a_MPP_Cluster-NoLicense-9.5.3.26.5-redhat7-x86_64.tar.bz2
五、测试方法
1).源集群和目标集群IP地址不同，节点数，分片，备份相同测试
1.ip信息
源集群IP地址：192.168.18.11，192.168.18.13，192.168.18.14
目标集群ip：192.168.18.31，192.168.18.33，192.168.18.34

2、创建测试表，并插入测试数据
3、源集群备份：对源集群做全备，增备测试

[gbase@v952man1 ~]$ gcadmin switchmode readonly

========== switch cluster mode...
 switch pre mode:                 [NORMAL]
 switch mode to                   [READONLY]
 switch after mode:                 [READONLY]
[gbase@v952man1 ~]$ gcrcman.py -d /home/gbase/8a-autobackup -e 'backup level 1 '  -P gbase@123 -p gbase@123 
01.31 00:44:21  check cluster topology begin 
01.31 00:44:21  node (192.168.18.11)  check topology begin 
01.31 00:44:23  node (192.168.18.11)  check topology success 
01.31 00:44:23  check cluster topology end 
01.31 00:44:23  BackUp start 
--------------------------------------------
01.31 00:44:23  node (192.168.18.11)  backup begin 
01.31 00:44:23  node (192.168.18.13)  backup begin 
01.31 00:44:23  node (192.168.18.14)  backup begin 
01.31 00:44:56  node (192.168.18.11)  backup success 
01.31 00:44:56  node (192.168.18.13)  backup success 
01.31 00:44:56  node (192.168.18.14)  backup success 
--------------------------------------------
01.31 00:44:56  BackUp end 
[gbase@v952man1 ~]$ gcadmin switchmode normal
 

4、将备份拷贝到目标集群，并修改属主：属组为gbase
5、全备恢复尝试(成功）
[gbase@gbaseman gcinstall]$ gcrcman.py -d /home/gbase/8a-autobackup -e 'show backup' -P gbase@123 -p gbase@123 
cycle    point    level    time
0    0    0    2023-01-30 22:58:16
[gbase@gbaseman gcinstall]$ gcadmin switchmode recovery

========== switch cluster mode...
 switch pre mode:                 [NORMAL]
 switch mode to                   [RECOVERY]
 switch after mode:                 [RECOVERY]
[gbase@gbaseman gcinstall]$ gcrcman.py -d /home/gbase/8a-autobackup -e 'recover 0' -P gbase@123 -p gbase@123 
01.31 00:18:35  check cluster topology begin 
01.31 00:18:35  node (192.168.18.31)  check topology begin 
01.31 00:18:37  node (192.168.18.31)  check topology success 
01.31 00:18:37  check cluster topology end 
01.31 00:18:37  check BackUp start 
--------------------------------------------
01.31 00:18:37  node (192.168.18.31)  check backup begin 
01.31 00:18:37  node (192.168.18.33)  check backup begin 
01.31 00:18:37  node (192.168.18.34)  check backup begin 
01.31 00:18:47  node (192.168.18.31)  check backup success 
01.31 00:18:47  node (192.168.18.33)  check backup success 
01.31 00:18:47  node (192.168.18.34)  check backup success 
--------------------------------------------
01.31 00:18:47  check BackUp success 
01.31 00:18:47  Recover start 
--------------------------------------------
01.31 00:18:48  node (192.168.18.31)  Recover begin 
01.31 00:18:48  node (192.168.18.33)  Recover begin 
01.31 00:18:48  node (192.168.18.34)  Recover begin 
01.31 00:19:03  node (192.168.18.31)  Recover success 
01.31 00:19:03  node (192.168.18.33)  Recover success 
01.31 00:19:03  node (192.168.18.34)  Recover success 
--------------------------------------------
01.31 00:19:04  Recover success AL]
[gbase@v952man1 ~]$ gcadmin switchmode normal

6.增备尝试恢复
[gbase@gbaseman ~]$ gcrcman.py -d /home/gbase/8a-autobackup -e 'show backup' -P gbase@123 -p gbase@123 
cycle    point    level    time
0    0    0    2023-01-30 22:58:16
0    1    1    2023-01-31 00:44:23
[gbase@gbaseman ~]$ gcadmin switchmode recovery

========== switch cluster mode...
 switch pre mode:                 [RECOVERY]
 switch mode to                   [RECOVERY]
 switch after mode:                 [RECOVERY]
[gbase@gbaseman ~]$ gcrcman.py -d /home/gbase/8a-autobackup -e 'recover 0 1' -P gbase@123 -p gbase@123 
01.31 01:12:34  check cluster topology begin 
01.31 01:12:34  node (192.168.18.31)  check topology begin 
01.31 01:12:37  node (192.168.18.31)  check topology success 
01.31 01:12:37  check cluster topology end 
01.31 01:12:37  check BackUp start 
--------------------------------------------
01.31 01:12:37  node (192.168.18.31)  check backup begin 
01.31 01:12:37  node (192.168.18.33)  check backup begin 
01.31 01:12:37  node (192.168.18.34)  check backup begin 
01.31 01:12:51  node (192.168.18.31)  check backup success 
01.31 01:12:51  node (192.168.18.33)  check backup success 
01.31 01:12:51  node (192.168.18.34)  check backup success 
--------------------------------------------
01.31 01:12:51  check BackUp success 
01.31 01:12:51  Recover start 
--------------------------------------------
01.31 01:12:52  node (192.168.18.31)  Recover begin 
01.31 01:12:52  node (192.168.18.33)  Recover begin 
01.31 01:12:52  node (192.168.18.34)  Recover begin 
01.31 01:13:07  node (192.168.18.31)  Recover success 
01.31 01:13:07  node (192.168.18.33)  Recover success 
01.31 01:13:07  node (192.168.18.34)  Recover success 
--------------------------------------------
01.31 01:13:08  Recover success 
[gbase@gbaseman ~]$ gcadmin switchmode normal

========== switch cluster mode...
 switch pre mode:                 [RECOVERY]
 switch mode to                   [NORMAL]
 switch after mode:                 [NORMAL]

2).源集群和目标集群ip地址相同，管理节点和数据节点数和分片数一致，distribution id不同
1.ip信息
源集群IP地址：192.168.18.11，192.168.18.13，192.168.18.14
目标集群ip：192.168.18.11，192.168.18.13，192.168.18.14
2.测试结果：成功
[gbase@gbaseman gcinstall]$ gcrcman.py -d /home/gbase/8a-autobackup -e 'show backup' -P gbase@123 -p gbase@123 
cycle    point    level    time
0    0    0    2023-01-30 22:58:16
[gbase@gbaseman gcinstall]$ gcadmin switchmode recovery

========== switch cluster mode...
 switch pre mode:                 [NORMAL]
 switch mode to                   [RECOVERY]
 switch after mode:                 [RECOVERY]
[gbase@gbaseman gcinstall]$ gcrcman.py -d /home/gbase/8a-autobackup -e 'recover 0' -P gbase@123 -p gbase@123 
01.30 23:58:04  check cluster topology begin 
01.30 23:58:04  node (192.168.18.11)  check topology begin 
01.30 23:58:06  node (192.168.18.11)  check topology success 
01.30 23:58:06  check cluster topology end 
01.30 23:58:06  check BackUp start 
--------------------------------------------
01.30 23:58:06  node (192.168.18.11)  check backup begin 
01.30 23:58:06  node (192.168.18.13)  check backup begin 
01.30 23:58:06  node (192.168.18.14)  check backup begin 
01.30 23:58:16  node (192.168.18.11)  check backup success 
01.30 23:58:16  node (192.168.18.13)  check backup success 
01.30 23:58:16  node (192.168.18.14)  check backup success 
--------------------------------------------
01.30 23:58:16  check BackUp success 
01.30 23:58:16  Recover start 
--------------------------------------------
01.30 23:58:17  node (192.168.18.11)  Recover begin 
01.30 23:58:17  node (192.168.18.13)  Recover begin 
01.30 23:58:17  node (192.168.18.14)  Recover begin 
01.30 23:58:32  node (192.168.18.11)  Recover success 
01.30 23:58:32  node (192.168.18.13)  Recover success 
01.30 23:58:32  node (192.168.18.14)  Recover success 
--------------------------------------------
01.30 23:58:33  Recover success

六、结论
根据测试得出结论，gcrcman跨集群备份恢复与ip地址无关与distribution id无关，恢复时集群的拓扑结构相同，包括管理节点、数据节点数及分片数需一致。并且备份需拷贝到与源集群对应的目标节点。