GaussDB T 0 1级备份 类似rman

最新推荐文章于 2024-05-27 11:01:23 发布
最新推荐文章于 2024-05-27 11:01:23 发布
阅读量401 收藏
点赞数
文章标签： 数据库 sql
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.csdn.net/jnrjian/article/details/127785713
版权

 对于使用zsql来进行备份和恢复；再进行restore恢复时，正确的姿势是直接使用增量备份进行恢复；而不是指定全备；然后再去应用增量，这也会报错。

[omm@gdb bin]$ python zctl.py -t stop
Successfully stopped instance.
[omm@gdb bin]$ python zctl.py -t start -m nomount
starting instance(nomount)
instance started
Please check log '/gdb/gaussdb/data/log/zenithstatus.log' or '/gdb/gaussdb/data/log/run/zengine.rlog' to find more fail reason.
[omm@gdb bin]$  rm -rf /gdb/gaussdb/data/data/* 
[omm@gdb bin]$  rm -rf /gdb/gaussdb/data/archive_log/* 
[omm@gdb bin]$ zsql / as sysdba

 
SQL> RESTORE DATABASE FROM '/tmp/roger_level0_02.bak';

Succeed.

SQL>  recover database;

 alter database open;
Succeed.

SQL> SQL> 

Succeed.

SQL> 
SQL> 
 

这是关于Gaussdb 100的物理差异增量备份恢复的例子；通过该测试可以窥测一些相关机制；供大家参加学习。

–开始归档

 

----归档
 python zctl.py -t stop
 python zctl.py -t start -m mount
 

alter database gauss noarchivelog;
 alter database gauss archivelog;
 

 

–差异增量0 级备份（压缩）
 

 BACKUP DATABASE INCREMENTAL LEVEL 0  FORMAT '/tmp/roger_level0_01.bak' as compressed backupset;

–差异增量1 级备份（压缩）
  BACKUP DATABASE INCREMENTAL LEVEL 1  FORMAT '/tmp/roger_level0_02.bak' as compressed backupset;

 

 
 

–查看备份进程情况
 

 

SQL> select * from V$BACKUP_PROCESS; TYPE PROGRESS STAGE STATUS ERR_NO ERR_MSG TOTAL_PROC FREE_PROC ---------- ------------ ---------- ---------- ------------ ---------- ------------ ------------ backup 64 data files processing 0 4 3 1 rows fetched. SQL> select * from V$BACKUP_PROCESS; TYPE PROGRESS STAGE STATUS ERR_NO ERR_MSG TOTAL_PROC FREE_PROC ---------- ------------ ---------- ---------- ------------ ---------- ------------ ------------ backup 100 end success 0 0 0 1 rows fetched. 
 

 
 

这里比较奇怪的是以为指定是备份集文件名字，其实是会创建该名称的目录，然后在目录下存放备份文件；可以看到是针对每个文件进行了备份。
 

—继续进行1级增量备份
 

 

 
 

+++ 恢复测试
 

 

[roger@mysqldb gaussdata]$ rm -rf user* [roger@mysqldb gaussdata]$ rm -rf sys* [roger@mysqldb gaussdata]$ rm -rf undo* [roger@mysqldb gaussdata]$ rm -rf temp* [roger@mysqldb gaussdata]$ zsql / as sysdba -q connected. SQL> RESTORE DATABASE FROM '/tmp/gaussdb_backup/roger_level0_01.bak'; GS-00855, restore failed, Failed to create the file /opt/gauss/gaussdata/data/cntl1, the error code was 17 
 

提示控制文件已经存在。在nomount下，居然不能覆盖控制文件。。。。好吧，我rm掉。
 

 

[roger@mysqldb gaussdata]$ rm cntl* [roger@mysqldb gaussdata]$ cd data/ [roger@mysqldb data]$ ls -ltr total 30720 -rw------- 1 roger roger 10485760 Feb 14 18:09 cntl1 -rw------- 1 roger roger 10485760 Feb 14 18:09 cntl2 -rw------- 1 roger roger 10485760 Feb 14 18:09 cntl3 [roger@mysqldb data]$ rm -rf * [roger@mysqldb gaussdata]$ zengine nomount -D /opt/gauss/gaussdata & [1] 91990 [roger@mysqldb gaussdata]$ starting instance(nomount) instance started 
 

–继续恢复测试
 

 

[roger@mysqldb data]$ zsql / as sysdba -q connected. SQL> restore database from '/tmp/gaussdb_backup/roger_level0_01.bak'; GS-00855, restore failed, Failed to create the file /opt/gauss/gaussdata/log1, the error code was 17 SQL> 
 

好吧。restore 跟log居然也有关系。好吧。 重新删干净后再次运行。
 

 

[roger@mysqldb gaussdata]$ zsql / as sysdba -q connected.
 

 

 
 

RESTORE DATABASE FROM '/tmp/roger_level0_01.bak';
 RESTORE DATABASE FROM '/tmp/roger_level0_02.bak';
 


  recover database;
 

 alter database open;
 

 

可以看到全备恢复成功了。
 

此时的全备restore的log信息如下：
 

 

UTC+8 2020-02-14 18:17:06.398|ZENGINE|00053|92312|INFO>[BACKUP] restore start, device:0, policy:, paral count 4, path:/tmp/gaussdb_backup/roger_level0_01.bak [knl_restore.c:1947] UTC+8 2020-02-14 18:17:06.398|ZENGINE|00053|92312|INFO>[ARCH] Already initialized [knl_archive.c:803] UTC+8 2020-02-14 18:17:06.403|ZENGINE|00000|95864|INFO>[BACKUP] parallel process 2 start [knl_bak_paral.c:996] UTC+8 2020-02-14 18:17:06.403|ZENGINE|00000|95865|INFO>[BACKUP] parallel process 3 start [knl_bak_paral.c:996] UTC+8 2020-02-14 18:17:06.403|ZENGINE|00000|95866|INFO>[BACKUP] parallel process 4 start [knl_bak_paral.c:996] UTC+8 2020-02-14 18:17:06.403|ZENGINE|00000|95863|INFO>[BACKUP] parallel process 1 start [knl_bak_paral.c:996] UTC+8 2020-02-14 18:17:06.532|ZENGINE|00053|94871532693656|INFO>[RESTORE] restore data_0_0, recent filesize 0K, section [0K, 16528K], bakcup id 1 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:17:06.634|ZENGINE|00053|94871532693656|INFO>[RESTORE] restore data_1_0, recent filesize 0K, section [0K, 16K], bakcup id 2 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:17:06.668|ZENGINE|00000|139818365449847|INFO>[RESTORE] fill datafile tail, start 16472K, end 16528K, backup id 1 [knl_bak_paral.c:763] UTC+8 2020-02-14 18:17:06.733|ZENGINE|00053|94871532693656|INFO>[RESTORE] restore data_3_0, recent filesize 0K, section [0K, 393216K], bakcup id 3 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:17:06.733|ZENGINE|00053|94871532693656|INFO>[RESTORE] restore data_3_1, recent filesize 0K, section [393216K, 531920K], bakcup id 4 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:17:06.947|ZENGINE|00053|94871532693656|INFO>[RESTORE] restore data_4_0, recent filesize 0K, section [0K, 204240K], bakcup id 5 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:17:07.261|ZENGINE|00053|94871532693656|INFO>[RESTORE] restore data_9_0, recent filesize 0K, section [0K, 16K], bakcup id 6 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:17:07.588|ZENGINE|00053|94871532693656|INFO>[RESTORE] restore data_11_0, recent filesize 0K, section [0K, 16K], bakcup id 7 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:17:07.892|ZENGINE|00053|94871532693656|INFO>[RESTORE] restore data_12_0, recent filesize 0K, section [0K, 89648K], bakcup id 8 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:17:08.586|ZENGINE|00000|139818365449849|INFO>[RESTORE] fill datafile tail, start 204192K, end 204240K, backup id 5 [knl_bak_paral.c:763] UTC+8 2020-02-14 18:17:09.043|ZENGINE|00000|95862|INFO>[RESTORE] restore datafiles finished [knl_restore.c:949] UTC+8 2020-02-14 18:17:09.043|ZENGINE|00000|95862|INFO>[BACKUP] create logfiles completed [knl_restore.c:565] UTC+8 2020-02-14 18:17:09.052|ZENGINE|00000|139818365449847|INFO>[RESTORE] start restore arch_12_0, backup id 9, log name /tmp/gaussdb_backup/roger_level0_01.bak/arch_12_0.bak [knl_bak_paral.c:673] UTC+8 2020-02-14 18:17:09.054|ZENGINE|00000|139818365449847|INFO>[ARCH] Record archive log file /opt/gauss/gaussdata/archive_log/arch_0_12.arc for log [0-12] start 0 end 1 [knl_archive.c:284] UTC+8 2020-02-14 18:17:09.146|ZENGINE|00000|95862|INFO>[RESTORE] restore logfiles finished, current backupset /tmp/gaussdb_backup/roger_level0_01.bak [knl_restore.c:964] UTC+8 2020-02-14 18:17:42.642|ZENGINE|00000|95862|INFO>[ARCH] Record archive log file /opt/gauss/gaussdata/archive_log/arch_0_13.arc for log [0-13] start 0 end 2 [knl_archive.c:284] UTC+8 2020-02-14 18:17:42.665|ZENGINE|00000|95863|INFO>[BACKUP] parallel process 1 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 18:17:42.672|ZENGINE|00000|95864|INFO>[BACKUP] parallel process 2 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 18:17:42.683|ZENGINE|00000|95865|INFO>[BACKUP] parallel process 3 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 18:17:42.685|ZENGINE|00000|95866|INFO>[BACKUP] parallel process 4 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 18:17:42.685|ZENGINE|00053|4995142158903240856|INFO>[BACKUP] restore success [knl_bak_common.c:1762] 
 

可以看到默认情况下即会启动4个进程进行恢复操作。同时我们可以看到restore进程在完成数据文件restore之后，还自动注册了归档。 应用了archive log 到最新
 

+++进行数据恢复操作
 

 

SQL> recover database; Succeed. SQL> alter database open; Succeed. SQL> conn roger/Roger007@127.0.0.1:1611 connected. SQL> select * from test; A B ---------------------------------------- -------------------- 1 www.enmotech.com 1 www.killdb.com 100 www.modb.pro 3 rows fetched. SQL> 
 

进一步观察此时recover的操作日志：
 

 

UTC+8 2020-02-14 18:21:00.863|ZENGINE|00053|92312|INFO>sucessfully alter database MOUNT [knl_database.c:1103] UTC+8 2020-02-14 18:21:00.899|ZENGINE|00053|92312|INFO>database start recovery [knl_recovery.c:1457] UTC+8 2020-02-14 18:21:00.899|ZENGINE|00053|92312|INFO>recovery from file:12,point:112,lfn:27160 [knl_recovery.c:1459] UTC+8 2020-02-14 18:21:00.899|ZENGINE|00053|92312|INFO>recovery expected least end with file:12,point:112,lfn:27160 [knl_recovery.c:1461] UTC+8 2020-02-14 18:21:01.010|ZENGINE|00053|139818365446296|INFO>[RCY] recovery real end with file:14,point:1,lfn:27247 [knl_recovery.c:1423] UTC+8 2020-02-14 18:21:01.010|ZENGINE|00053|139818365446296|INFO>[RCY] current lfn 27247, rcy point lfn 27160, consistent point 27160, lrp point lfn 27160 [knl_recovery.c:1425] UTC+8 2020-02-14 18:21:16.990|ZENGINE|00053|206158522520|INFO>start to alter database OPEN [knl_database.c:1322] UTC+8 2020-02-14 18:21:17.071|ZENGINE|00053|139818365446296|INFO>[ARCH] Start ARCH thread for ARCHIVE_DEST_1[/opt/gauss/gaussdata/archive_log] [knl_archive.c:881] UTC+8 2020-02-14 18:21:17.071|ZENGINE|00053|15214422880117942424|INFO>The last shutdown is inconsistent [knl_database.c:1379] UTC+8 2020-02-14 18:21:17.071|ZENGINE|00053|92312|INFO>database start recovery [knl_recovery.c:1457] UTC+8 2020-02-14 18:21:17.071|ZENGINE|00053|92312|INFO>recovery from file:13,point:750,lfn:27247 [knl_recovery.c:1459] UTC+8 2020-02-14 18:21:17.071|ZENGINE|00053|92312|INFO>recovery expected least end with file:13,point:750,lfn:27247 [knl_recovery.c:1461] UTC+8 2020-02-14 18:21:17.082|ZENGINE|00053|139818365446296|INFO>[RCY] recovery real end with file:14,point:1,lfn:27247 [knl_recovery.c:1423] UTC+8 2020-02-14 18:21:17.082|ZENGINE|00053|139818365446296|INFO>[RCY] current lfn 27247, rcy point lfn 27247, consistent point 27247, lrp point lfn 27247 [knl_recovery.c:1425] UTC+8 2020-02-14 18:21:17.148|ZENGINE|00053|139822660413592|INFO>no valid standby configuration [knl_lns.c:1727] UTC+8 2020-02-14 18:21:17.152|ZENGINE|00053|92312|INFO>[DB] sse42 available 1 [knl_database.c:1440] UTC+8 2020-02-14 18:21:17.152|ZENGINE|00053|92312|INFO>sucessfully alter database OPEN [knl_database.c:1441] UTC+8 2020-02-14 18:21:17.300|ZENGINE|00000|92000|INFO>rollback thread closed [knl_tran.c:1203] UTC+8 2020-02-14 18:21:17.304|ZENGINE|00000|91999|INFO>rollback thread closed [knl_tran.c:1203] UTC+8 2020-02-14 18:21:17.924|ZENGINE|00054|99048|INFO>load index parts: load index parts, uid: 0, tid: 278, iid: 0, index partcnt: 4032 [knl_dc_part.c:993] UTC+8 2020-02-14 18:21:17.926|ZENGINE|00054|99048|INFO>load index parts: load index parts, uid: 0, tid: 302, iid: 0, index partcnt: 4032 [knl_dc_part.c:993] UTC+8 2020-02-14 18:21:17.928|ZENGINE|00054|99048|INFO>load index parts: load index parts, uid: 0, tid: 274, iid: 0, index partcnt: 4032 [knl_dc_part.c:993] UTC+8 2020-02-14 18:21:17.933|ZENGINE|00054|99048|INFO>load index parts: load index parts, uid: 0, tid: 276, iid: 0, index partcnt: 4032 [knl_dc_part.c:993] 
 

可以看到recover的起点：lfn:27160；recover的终点：lfn 27247
 

++++ 第二次恢复测试
 

 

删掉所有数据文件，redo文件，控制文件以及归档文件。 [roger@mysqldb gaussdata]$ rm -rf log4 log3 log2 log5 log6 log1 [roger@mysqldb gaussdata]$ rm -rf sys* [roger@mysqldb gaussdata]$ rm -rf undo [roger@mysqldb gaussdata]$ rm -rf temp* [roger@mysqldb gaussdata]$ cd data/ [roger@mysqldb data]$ ls -ltr total 30720 -rw------- 1 roger roger 10485760 Feb 14 18:51 cntl1 -rw------- 1 roger roger 10485760 Feb 14 18:51 cntl2 -rw------- 1 roger roger 10485760 Feb 14 18:51 cntl3 [roger@mysqldb data]$ rm -rf * [roger@mysqldb gaussdata]$ cd archive_log/ [roger@mysqldb archive_log]$ ls -ltr total 1335616 -rw------- 1 roger roger 133393920 Feb 14 12:55 arch_0_1.arc -rw------- 1 roger roger 133582848 Feb 14 12:55 arch_0_2.arc -rw------- 1 roger roger 133707776 Feb 14 12:55 arch_0_3.arc -rw------- 1 roger roger 133812736 Feb 14 12:56 arch_0_4.arc -rw------- 1 roger roger 132929536 Feb 14 16:32 arch_0_5.arc -rw------- 1 roger roger 134062592 Feb 14 16:32 arch_0_6.arc -rw------- 1 roger roger 132893184 Feb 14 16:32 arch_0_7.arc -rw------- 1 roger roger 134105088 Feb 14 16:39 arch_0_8.arc -rw------- 1 roger roger 134017536 Feb 14 16:48 arch_0_9.arc -rw------- 1 roger roger 134163456 Feb 14 16:48 arch_0_10.arc -rw------- 1 roger roger 30539264 Feb 14 17:53 arch_0_11.arc -rw------- 1 roger roger 57344 Feb 14 17:58 arch_0_12.arc -rw------- 1 roger roger 384000 Feb 14 18:06 arch_0_13.arc [roger@mysqldb archive_log]$ rm -rf * [roger@mysqldb archive_log]$ ls [roger@mysqldb archive_log]$ cd .. [roger@mysqldb gaussdata]$ ls -ltr total 5083144 drwx------ 2 roger roger 42 Feb 13 16:23 dbs -rw------- 1 roger roger 0 Feb 13 16:23 zengine.lck drwx------ 5 roger roger 87 Feb 13 16:39 log drwx------ 2 roger roger 4096 Feb 13 21:50 trc -rw------- 1 roger roger 134217728 Feb 14 16:48 log3_bak -rw------- 1 roger roger 134217728 Feb 14 16:48 log4_bak -rw------- 1 roger roger 134217728 Feb 14 17:53 log5_bak -rw------- 1 roger roger 134217728 Feb 14 17:58 log6_bak -rw------- 1 roger roger 134217728 Feb 14 18:06 log1_bak -rw------- 1 roger roger 134217728 Feb 14 18:06 log2_bak drwx------ 2 roger roger 85 Feb 14 18:12 protect -rw------- 1 roger roger 104857600 Feb 14 18:17 user5 -rw------- 1 roger roger 1073741824 Feb 14 18:17 user2 -rw------- 1 roger roger 1073741824 Feb 14 18:17 user3 -rw------- 1 roger roger 1073741824 Feb 14 18:17 user4 drwx------ 2 roger roger 62 Feb 14 18:20 cfg -rw------- 1 roger roger 1073741824 Feb 14 18:26 user1 drwx------ 2 roger roger 6 Feb 14 18:57 data drwx------ 2 roger roger 6 Feb 14 18:57 archive_log [roger@mysqldb gaussdata]$ rm -rf user* [roger@mysqldb gaussdata]$ rm -rf log* [roger@mysqldb gaussdata]$ ls -ltr total 4 drwx------ 2 roger roger 42 Feb 13 16:23 dbs -rw------- 1 roger roger 0 Feb 13 16:23 zengine.lck drwx------ 2 roger roger 4096 Feb 13 21:50 trc drwx------ 2 roger roger 85 Feb 14 18:12 protect drwx------ 2 roger roger 62 Feb 14 18:20 cfg drwx------ 2 roger roger 6 Feb 14 18:57 data drwx------ 2 roger roger 6 Feb 14 18:57 archive_log [roger@mysqldb gaussdata]$ mkdir log 
 

–启动实例到nomount
 

 

[roger@mysqldb gaussdata]$ zengine nomount -D /opt/gauss/gaussdata & [1] 127423 [roger@mysqldb gaussdata]$ starting instance(nomount) instance started [roger@mysqldb gaussdata]$ [roger@mysqldb gaussdata]$ zsql / as sysdba -q connected. 
 

–进行restore & recover
 

 

 rm -rf /gdb/gaussdb/data/data/*                 */
  rm -rf /gdb/gaussdb/data/archive_log/*                */
 

 

 

 

 

SQL> restore database from '/tmp/gaussdb_backup/roger_level0_01.bak'; Succeed. SQL> recover database; Succeed. SQL> alter database open; Succeed. SQL> SQL> conn roger/Roger007@127.0.0.1:1611 connected. SQL> select * from test; A B ---------------------------------------- -------------------- 1 www.enmotech.com 1 www.killdb.com 2 rows fetched. ----丢失了最新的一条数据。 
 

—如下是相关日志
 

 

TC+8 2020-02-14 18:58:54.292|ZENGINE|00053|127593|INFO>[BACKUP] restore start, device:0, policy:, paral count 4, path:/tmp/gaussdb_backup/roger_level0_01.bak [knl_restore.c:1947] UTC+8 2020-02-14 18:58:54.292|ZENGINE|00053|127593|INFO>[ARCH] Initialization complete [knl_archive.c:824] UTC+8 2020-02-14 18:58:54.304|ZENGINE|00000|127988|INFO>[BACKUP] parallel process 1 start [knl_bak_paral.c:996] UTC+8 2020-02-14 18:58:54.316|ZENGINE|00000|127989|INFO>[BACKUP] parallel process 2 start [knl_bak_paral.c:996] UTC+8 2020-02-14 18:58:54.316|ZENGINE|00000|127990|INFO>[BACKUP] parallel process 3 start [knl_bak_paral.c:996] UTC+8 2020-02-14 18:58:54.316|ZENGINE|00000|127991|INFO>[BACKUP] parallel process 4 start [knl_bak_paral.c:996] UTC+8 2020-02-14 18:58:54.420|ZENGINE|00053|93849330512489|INFO>[RESTORE] restore data_0_0, recent filesize 0K, section [0K, 16528K], bakcup id 1 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:58:54.520|ZENGINE|00053|93849330512489|INFO>[RESTORE] restore data_1_0, recent filesize 0K, section [0K, 16K], bakcup id 2 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:58:54.576|ZENGINE|00000|139693811430388|INFO>[RESTORE] fill datafile tail, start 16472K, end 16528K, backup id 1 [knl_bak_paral.c:763] UTC+8 2020-02-14 18:58:54.621|ZENGINE|00053|93849330512489|INFO>[RESTORE] restore data_3_0, recent filesize 0K, section [0K, 393216K], bakcup id 3 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:58:54.621|ZENGINE|00053|93849330512489|INFO>[RESTORE] restore data_3_1, recent filesize 0K, section [393216K, 531920K], bakcup id 4 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:58:54.843|ZENGINE|00053|93849330512489|INFO>[RESTORE] restore data_4_0, recent filesize 0K, section [0K, 204240K], bakcup id 5 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:58:55.156|ZENGINE|00053|93849330512489|INFO>[RESTORE] restore data_9_0, recent filesize 0K, section [0K, 16K], bakcup id 6 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:58:55.477|ZENGINE|00053|93849330512489|INFO>[RESTORE] restore data_11_0, recent filesize 0K, section [0K, 16K], bakcup id 7 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:58:55.806|ZENGINE|00053|93849330512489|INFO>[RESTORE] restore data_12_0, recent filesize 0K, section [0K, 89648K], bakcup id 8 [knl_bak_paral.c:374] UTC+8 2020-02-14 18:58:57.431|ZENGINE|00000|139693811430390|INFO>[RESTORE] fill datafile tail, start 204192K, end 204240K, backup id 5 [knl_bak_paral.c:763] UTC+8 2020-02-14 18:58:58.019|ZENGINE|00000|127987|INFO>[RESTORE] restore datafiles finished [knl_restore.c:949] UTC+8 2020-02-14 18:58:58.019|ZENGINE|00000|127987|INFO>[BACKUP] create logfiles completed [knl_restore.c:565] UTC+8 2020-02-14 18:58:58.031|ZENGINE|00000|139693811430388|INFO>[RESTORE] start restore arch_12_0, backup id 9, log name /tmp/gaussdb_backup/roger_level0_01.bak/arch_12_0.bak [knl_bak_paral.c:673] UTC+8 2020-02-14 18:58:58.031|ZENGINE|00000|139693811430388|INFO>[BACKUP] Create /opt/gauss/gaussdata/archive_log/arch_0_12.arc [knl_restore.c:362] UTC+8 2020-02-14 18:58:58.035|ZENGINE|00000|139693811430388|INFO>[ARCH] Record archive log file /opt/gauss/gaussdata/archive_log/arch_0_12.arc for log [0-12] start 0 end 1 [knl_archive.c:284] UTC+8 2020-02-14 18:58:58.122|ZENGINE|00000|127987|INFO>[RESTORE] restore logfiles finished, current backupset /tmp/gaussdb_backup/roger_level0_01.bak [knl_restore.c:964] UTC+8 2020-02-14 18:59:19.783|ZENGINE|00000|127988|INFO>[BACKUP] parallel process 1 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 18:59:19.785|ZENGINE|00000|127989|INFO>[BACKUP] parallel process 2 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 18:59:19.792|ZENGINE|00000|127990|INFO>[BACKUP] parallel process 3 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 18:59:19.802|ZENGINE|00000|127991|INFO>[BACKUP] parallel process 4 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 18:59:19.804|ZENGINE|00053|4995142158903276137|INFO>[BACKUP] restore success [knl_bak_common.c:1762] UTC+8 2020-02-14 18:59:40.516|ZENGINE|00053|7162257363678917225|INFO>[BACKUP] restore start, device:0, policy:, paral count 4, path:/tmp/gaussdb_backup/roger_level1_01.bak [knl_restore.c:1947] ....... UTC+8 2020-02-14 18:59:46.691|ZENGINE|00053|206158557801|INFO>start to alter database MOUNT [knl_database.c:1047] UTC+8 2020-02-14 18:59:46.745|ZENGINE|00053|127593|INFO>[ARCH] Already initialized [knl_archive.c:803] UTC+8 2020-02-14 18:59:46.745|ZENGINE|00053|127593|INFO>sucessfully alter database MOUNT [knl_database.c:1103] UTC+8 2020-02-14 18:59:46.762|ZENGINE|00053|127593|INFO>database start recovery [knl_recovery.c:1457] UTC+8 2020-02-14 18:59:46.762|ZENGINE|00053|127593|INFO>recovery from file:12,point:112,lfn:27160 [knl_recovery.c:1459] UTC+8 2020-02-14 18:59:46.762|ZENGINE|00053|127593|INFO>recovery expected least end with file:12,point:112,lfn:27160 [knl_recovery.c:1461] UTC+8 2020-02-14 18:59:46.807|ZENGINE|00053|139693811429993|INFO>[RCY] recovery real end with file:13,point:1,lfn:27160 [knl_recovery.c:1423] UTC+8 2020-02-14 18:59:46.807|ZENGINE|00053|139693811429993|INFO>[RCY] current lfn 27160, rcy point lfn 27160, consistent point 27160, lrp point lfn 27160 [knl_recovery.c:1425] UTC+8 2020-02-14 19:00:06.205|ZENGINE|00053|206158557801|INFO>start to alter database OPEN [knl_database.c:1322] UTC+8 2020-02-14 19:00:06.286|ZENGINE|00053|139693811429993|INFO>[ARCH] Start ARCH thread for ARCHIVE_DEST_1[/opt/gauss/gaussdata/archive_log] [knl_archive.c:881] UTC+8 2020-02-14 19:00:06.286|ZENGINE|00053|8527087606868079209|INFO>The last shutdown is inconsistent [knl_database.c:1379] UTC+8 2020-02-14 19:00:06.286|ZENGINE|00053|127593|INFO>database start recovery [knl_recovery.c:1457] UTC+8 2020-02-14 19:00:06.286|ZENGINE|00053|127593|INFO>recovery from file:12,point:112,lfn:27160 [knl_recovery.c:1459] UTC+8 2020-02-14 19:00:06.286|ZENGINE|00053|127593|INFO>recovery expected least end with file:12,point:112,lfn:27160 [knl_recovery.c:1461] UTC+8 2020-02-14 19:00:06.296|ZENGINE|00053|139693811429993|INFO>[RCY] recovery real end with file:13,point:1,lfn:27160 [knl_recovery.c:1423] UTC+8 2020-02-14 19:00:06.296|ZENGINE|00053|139693811429993|INFO>[RCY] current lfn 27160, rcy point lfn 27160, consistent point 27160, lrp point lfn 27160 [knl_recovery.c:1425] UTC+8 2020-02-14 19:00:06.386|ZENGINE|00053|139693811429993|INFO>no valid standby configuration [knl_lns.c:1727] UTC+8 2020-02-14 19:00:06.388|ZENGINE|00053|127593|INFO>[DB] sse42 available 1 [knl_database.c:1440] UTC+8 2020-02-14 19:00:06.388|ZENGINE|00053|127593|INFO>sucessfully alter database OPEN [knl_database.c:1441] UTC+8 2020-02-14 19:00:06.495|ZENGINE|00000|127432|INFO>rollback thread closed [knl_tran.c:1203] UTC+8 2020-02-14 19:00:06.495|ZENGINE|00000|127433|INFO>rollback thread closed [knl_tran.c:1203] UTC+8 2020-02-14 19:00:06.897|ZENGINE|00054|128941|INFO>load index parts: load index parts, uid: 0, tid: 278, iid: 0, index partcnt: 4032 [knl_dc_part.c:993] 
 

我们可以看到此时恢复到的点居然是：onsistent point 27160 这比之前归档存在的情况下恢复的点要小的多；缺乏严格的校验机制。因此最后一条增量数据丢失。
 

++++ 最后一次恢复测试
 

–删掉所有文件（数据文件/temp/controlfile/redo log)
 

 

[roger@mysqldb gaussdata]$ rm -rf user* [roger@mysqldb gaussdata]$ rm -rf temp* [roger@mysqldb gaussdata]$ rm -rf sys* [roger@mysqldb gaussdata]$ rm -rf log1 log2 log3 log4 log5 log6 [roger@mysqldb gaussdata]$ rm -rf archive_log/* [roger@mysqldb gaussdata]$ rm -rf data/cntl* [roger@mysqldb gaussdata]$ rm -rf undo [roger@mysqldb gaussdata]$ ls -ltr total 4 drwx------ 2 roger roger 42 Feb 13 16:23 dbs -rw------- 1 roger roger 0 Feb 13 16:23 zengine.lck drwx------ 2 roger roger 4096 Feb 13 21:50 trc drwx------ 5 roger roger 64 Feb 14 18:58 log drwx------ 2 roger roger 62 Feb 14 18:59 cfg drwx------ 2 roger roger 85 Feb 14 19:18 protect drwx------ 2 roger roger 6 Feb 14 19:58 archive_log drwx------ 2 roger roger 6 Feb 14 19:58 data 
 

–启动实例到nomount
 

 

[roger@mysqldb gaussdata]$ zengine nomount -D /opt/gauss/gaussdata & [1] 44946 [roger@mysqldb gaussdata]$ starting instance(nomount) instance started 
 

–开始全库restore
 

 

[roger@mysqldb gaussdata]$ zsql / as sysdba -q connected. SQL> restore database from '/tmp/gaussdb_backup/roger_level0_01.bak'; Succeed. SQL> 
 

此时的日志信息如下所示：
 

 

UTC+8 2020-02-14 20:01:18.882|ZENGINE|00000|45522|INFO>agent thread started, tid:139927433565952, close:0 [srv_agent.c:359] UTC+8 2020-02-14 20:01:18.882|ZENGINE|00053|45522|INFO>[LOG] file '/opt/gauss/gaussdata/log/audit/zengine.aud' is added [cm_log.c:641] UTC+8 2020-02-14 20:01:26.671|ZENGINE|00053|45522|INFO>[BACKUP] restore start, device:0, policy:, paral count 4, path:/tmp/gaussdb_backup/roger_level0_01.bak [knl_restore.c:1947] UTC+8 2020-02-14 20:01:26.671|ZENGINE|00053|45522|INFO>[ARCH] Initialization complete [knl_archive.c:824] UTC+8 2020-02-14 20:01:26.687|ZENGINE|00000|45608|INFO>[BACKUP] parallel process 4 start [knl_bak_paral.c:996] UTC+8 2020-02-14 20:01:26.698|ZENGINE|00000|45605|INFO>[BACKUP] parallel process 1 start [knl_bak_paral.c:996] UTC+8 2020-02-14 20:01:26.698|ZENGINE|00000|45606|INFO>[BACKUP] parallel process 2 start [knl_bak_paral.c:996] UTC+8 2020-02-14 20:01:26.698|ZENGINE|00000|45607|INFO>[BACKUP] parallel process 3 start [knl_bak_paral.c:996] UTC+8 2020-02-14 20:01:26.818|ZENGINE|00053|93948114678226|INFO>[RESTORE] restore data_0_0, recent filesize 0K, section [0K, 16528K], bakcup id 1 [knl_bak_paral.c:374] UTC+8 2020-02-14 20:01:26.919|ZENGINE|00053|93948114678226|INFO>[RESTORE] restore data_1_0, recent filesize 0K, section [0K, 16K], bakcup id 2 [knl_bak_paral.c:374] UTC+8 2020-02-14 20:01:26.982|ZENGINE|00000|139925739581989|INFO>[RESTORE] fill datafile tail, start 16472K, end 16528K, backup id 1 [knl_bak_paral.c:763] UTC+8 2020-02-14 20:01:27.019|ZENGINE|00053|93948114678226|INFO>[RESTORE] restore data_3_0, recent filesize 0K, section [0K, 393216K], bakcup id 3 [knl_bak_paral.c:374] UTC+8 2020-02-14 20:01:27.019|ZENGINE|00053|93948114678226|INFO>[RESTORE] restore data_3_1, recent filesize 0K, section [393216K, 531920K], bakcup id 4 [knl_bak_paral.c:374] UTC+8 2020-02-14 20:01:27.216|ZENGINE|00053|93948114678226|INFO>[RESTORE] restore data_4_0, recent filesize 0K, section [0K, 204240K], bakcup id 5 [knl_bak_paral.c:374] UTC+8 2020-02-14 20:01:27.541|ZENGINE|00053|93948114678226|INFO>[RESTORE] restore data_9_0, recent filesize 0K, section [0K, 16K], bakcup id 6 [knl_bak_paral.c:374] UTC+8 2020-02-14 20:01:27.859|ZENGINE|00053|93948114678226|INFO>[RESTORE] restore data_11_0, recent filesize 0K, section [0K, 16K], bakcup id 7 [knl_bak_paral.c:374] UTC+8 2020-02-14 20:01:28.167|ZENGINE|00053|93948114678226|INFO>[RESTORE] restore data_12_0, recent filesize 0K, section [0K, 89648K], bakcup id 8 [knl_bak_paral.c:374] UTC+8 2020-02-14 20:01:29.895|ZENGINE|00000|139925739581991|INFO>[RESTORE] fill datafile tail, start 204192K, end 204240K, backup id 5 [knl_bak_paral.c:763] UTC+8 2020-02-14 20:01:30.504|ZENGINE|00000|45604|INFO>[RESTORE] restore datafiles finished [knl_restore.c:949] UTC+8 2020-02-14 20:01:30.504|ZENGINE|00000|45604|INFO>[BACKUP] create logfiles completed [knl_restore.c:565] UTC+8 2020-02-14 20:01:30.515|ZENGINE|00000|139925739581989|INFO>[RESTORE] start restore arch_12_0, backup id 9, log name /tmp/gaussdb_backup/roger_level0_01.bak/arch_12_0.bak [knl_bak_paral.c:673] UTC+8 2020-02-14 20:01:30.515|ZENGINE|00000|139925739581989|INFO>[BACKUP] Create /opt/gauss/gaussdata/archive_log/arch_0_12.arc [knl_restore.c:362] UTC+8 2020-02-14 20:01:30.534|ZENGINE|00000|139925739581989|INFO>[ARCH] Record archive log file /opt/gauss/gaussdata/archive_log/arch_0_12.arc for log [0-12] start 0 end 1 [knl_archive.c:284] UTC+8 2020-02-14 20:01:30.610|ZENGINE|00000|45604|INFO>[RESTORE] restore logfiles finished, current backupset /tmp/gaussdb_backup/roger_level0_01.bak [knl_restore.c:964] UTC+8 2020-02-14 20:02:06.663|ZENGINE|00000|45605|INFO>[BACKUP] parallel process 1 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 20:02:06.673|ZENGINE|00000|45606|INFO>[BACKUP] parallel process 2 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 20:02:06.682|ZENGINE|00000|45607|INFO>[BACKUP] parallel process 3 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 20:02:06.693|ZENGINE|00000|45608|INFO>[BACKUP] parallel process 4 stop [knl_bak_paral.c:1022] UTC+8 2020-02-14 20:02:06.693|ZENGINE|00053|4995142158903194066|INFO>[BACKUP] restore success [knl_bak_common.c:1762] 
 

–开始recover
 

 

SQL> recover database; Succeed. 
 

此时日志信息如下：
 

 

UTC+8 2020-02-14 20:03:47.309|ZENGINE|00053|206158475730|INFO>start to alter database MOUNT [knl_database.c:1047] UTC+8 2020-02-14 20:03:47.386|ZENGINE|00053|45522|INFO>[ARCH] Already initialized [knl_archive.c:803] UTC+8 2020-02-14 20:03:47.386|ZENGINE|00053|45522|INFO>sucessfully alter database MOUNT [knl_database.c:1103] UTC+8 2020-02-14 20:03:47.445|ZENGINE|00053|45522|INFO>database start recovery [knl_recovery.c:1457] UTC+8 2020-02-14 20:03:47.445|ZENGINE|00053|45522|INFO>recovery from file:12,point:112,lfn:27160 [knl_recovery.c:1459] UTC+8 2020-02-14 20:03:47.445|ZENGINE|00053|45522|INFO>recovery expected least end with file:12,point:112,lfn:27160 [knl_recovery.c:1461] UTC+8 2020-02-14 20:03:47.491|ZENGINE|00053|139925739581906|INFO>[RCY] recovery real end with file:13,point:1,lfn:27160 [knl_recovery.c:1423] UTC+8 2020-02-14 20:03:47.491|ZENGINE|00053|139925739581906|INFO>[RCY] current lfn 27160, rcy point lfn 27160, consistent point 27160, lrp point lfn 27160 [knl_recovery.c:1425] 
 

可以看到recover操作会自动将数据库mount。
 

 

--dump controlfile SQL> alter system dump ctrlfile; Succeed. Elapsed: 0.014 sec 
 

此时控制文件dump信息如下：
 

 

[roger@mysqldb trc]$ cat ctrl_page.trc information of page 0-1 lsn: 0 pcn: 0 size: 16384 type: ctrl next_ext: 1023-0 core information: version: 1-0-1-0 startup times: 4 dbid times: 773766796 database name: ROGER init time: 2020-02-13 16:50:50 scn: 388345312780289 table$ entry: 2 ix_table$1 entry: 34 ix_table$2 entry: 42 column$ entry: 10 ix_column$ entry: 50 index$ entry: 18 ix_index$1 entry: 58 ix_index$2 entry: 66 user$_entry: 26 ix_user$1 entry: 74 ix_user$2 entry: 82 rcy point: 0-12-112-27160 lrp point: 0-12-112-27160 raft flush point: scn(0)-lfn(0)-raft_index(0) ckpt_id: 308 dw_start: 4 dw_end: 4 lsn: 8775411 lfn: 27160 build completed: 1 log count: 6 log hwm: 6 log first: 0 log last: 0 archive mode: 1 archive logs: 0-0-0-0-0-0-0-0-0-0 db_role: 0 protect mode: 0 space count: 7 device count: 13 page size: 8192 undo segments: 32 reset logs: 0-0-0 archived_start: 0 archived_end: 1 logic replication mode: 0 shutdown consistency: 0 open inconsistency: 0 consistent lfn: 27160 storage information: logfiles information: id name size hwm seq block_size flg type status forward backward #0 /opt/gauss/gaussdata/log1 134217728 0 0 512 0 1 3 0 0 #1 /opt/gauss/gaussdata/log2 134217728 0 0 512 0 1 1 0 0 #2 /opt/gauss/gaussdata/log3 134217728 0 0 512 0 1 1 0 0 #3 /opt/gauss/gaussdata/log4 134217728 0 0 512 0 1 1 0 0 #4 /opt/gauss/gaussdata/log5 134217728 0 0 512 0 1 1 0 0 #5 /opt/gauss/gaussdata/log6 134217728 0 0 512 0 1 1 0 0 spaces information: id spaceid used name flg block_size extent_size file_hwm org_scn files #0 0 1 SYSTEM 1 8192 8 1 6762045441 0 #1 1 1 TEMP 1 8192 16 2 9270042625 1, 2 #2 2 1 UNDO 1 8192 1 1 10840301569 3 #3 3 1 USERS 9 8192 8 5 16813305857 4, 5, 6, 7, 8 #4 4 1 TEMP2 9 8192 8 2 106346856449 9, 10 #5 5 1 TEMP2_UNDO 1 8192 1 1 108784082945 11 #6 6 1 SYSAUX 1 8192 8 1 110029156353 12 datafiles information: id dfileid used name size block_size flg type auto_extend_size auto_extend_maxsize archive log information: id recid dest_id rst_id asn stamp blocks block_size first last name #0 1 0 0 12 635025690515372 112 512 387183961239553 388346659434497 /opt/gauss/gaussdata/archive_log/arch_0_12.arc 
 

—再次重复前面的步骤进行恢复；并使用strace跟踪
 

 

[roger@mysqldb gaussdata]$ zengine nomount -D /opt/gauss/gaussdata & [1] 59647 [roger@mysqldb gaussdata]$ starting instance(nomount) instance started [roger@mysqldb gaussdata]$ [roger@mysqldb gaussdata]$ zsql / as sysdba -q connected. SQL> restore database from '/tmp/gaussdb_backup/roger_level0_01.bak'; Succeed. SQL> recover database; Succeed. SQL> alter database open; Succeed. SQL> [roger@mysqldb roger_level1_01.bak]$ strace -fr -F -o /tmp/gauss_recover.log -p 59647 Process 59647 attached with 17 threads Process 61870 attached Process 61871 attached Process 61872 attached Process 61873 attached Process 61874 attached Process 62876 attached Process 62916 attached Process 62915 attached ^CProcess 59647 detached Process 59648 detached Process 59649 detached Process 59650 detached Process 59651 detached Process 59652 detached Process 59653 detached Process 59654 detached Process 59655 detached Process 59658 detached Process 59659 detached Process 59660 detached Process 59661 detached Process 59662 detached Process 59888 detached Process 62876 detached [root@mysqldb tmp]# cat /tmp/gauss_recover.log |grep 'open' 59888 0.000236 open("/tmp/gaussdb_backup/roger_level0_01.bak/backupset", O_RDWR|O_SYNC) = 17 59888 0.000049 open("/opt/gauss/gaussdata/cfg/zengine.ini", O_RDONLY) = 18 59888 0.000062 open("/opt/gauss/gaussdata/cfg/zengine.ini_bak", O_RDWR|O_CREAT|O_TRUNC, 0600) = 19 59888 0.000074 open("/opt/gauss/gaussdata/cfg/zengine.ini_tmp", O_RDWR|O_CREAT|O_TRUNC|O_SYNC, 0600) = 18 59888 0.000044 open("/tmp/gaussdb_backup/roger_level0_01.bak/ctrl_0_0.bak", O_RDWR|O_SYNC) = 17 61870 0.000701 open("/opt/gauss/gaussdata/cfg/zengine.ini", O_RDONLY) = 18 61870 0.000610 open("/opt/gauss/gaussdata/cfg/zengine.ini_bak", O_RDWR|O_CREAT|O_TRUNC, 0600 <unfinished ...> 61870 0.000321 <... open resumed> ) = 19 61870 0.000038 open("/opt/gauss/gaussdata/cfg/zengine.ini_tmp", O_RDWR|O_CREAT|O_TRUNC|O_SYNC, 0600) = 18 61870 0.000046 open("/opt/gauss/gaussdata/data/cntl1", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 18 61870 0.000053 open("/opt/gauss/gaussdata/data/cntl2", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 19 61870 0.000044 open("/opt/gauss/gaussdata/data/cntl3", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 20 61870 0.000453 open("/opt/gauss/gaussdata/system", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 17 61870 0.000174 open("/opt/gauss/gaussdata/temp1_01", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 18 61870 0.000288 open("/opt/gauss/gaussdata/temp1_02", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 19 61870 0.000362 open("/opt/gauss/gaussdata/undo", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 20 61870 0.000314 open("/opt/gauss/gaussdata/user1", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 21 61870 0.000046 open("/opt/gauss/gaussdata/user2", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 22 61870 0.000032 open("/opt/gauss/gaussdata/user3", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 23 61870 0.000039 open("/opt/gauss/gaussdata/user4", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 24 61870 0.000025 open("/opt/gauss/gaussdata/user5", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 25 61870 0.000024 open("/opt/gauss/gaussdata/temp2_01", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 26 61870 0.000025 open("/opt/gauss/gaussdata/temp2_02", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 27 61870 0.000026 open("/opt/gauss/gaussdata/temp2_undo", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 28 61870 0.000044 open("/opt/gauss/gaussdata/sysaux", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 29 59888 0.000007 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_SYSTEM_0_0.bak", O_RDWR|O_SYNC <unfinished ...> 59888 0.000007 <... open resumed> ) = 30 61871 0.005724 open("/opt/gauss/gaussdata/system", O_RDWR|O_SYNC <unfinished ...> 61871 0.000268 <... open resumed> ) = 31 59888 0.000036 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_TEMP_1_0.bak", O_RDWR|O_SYNC) = 32 61872 0.000275 open("/opt/gauss/gaussdata/temp1_01", O_RDWR|O_SYNC <unfinished ...> 61872 0.000046 <... open resumed> ) = 33 59888 0.000058 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_UNDO_3_0.bak", O_RDWR|O_SYNC <unfinished ...> 59888 0.000040 <... open resumed> ) = 32 61872 0.000004 open("/opt/gauss/gaussdata/undo", O_RDWR|O_SYNC <unfinished ...> 61872 0.000705 <... open resumed> ) = 33 59888 0.000051 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_UNDO_3_1.bak", O_RDWR|O_SYNC) = 34 61873 0.001613 open("/opt/gauss/gaussdata/undo", O_RDWR|O_SYNC <unfinished ...> 61873 0.000354 <... open resumed> ) = 35 59888 0.000035 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_USERS_4_0.bak", O_RDWR|O_SYNC <unfinished ...> 59888 0.000053 <... open resumed> ) = 30 59888 0.000033 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_TEMP2_9_0.bak", O_RDWR|O_SYNC) = 31 61871 0.000476 open("/opt/gauss/gaussdata/temp2_01", O_RDWR|O_SYNC) = 36 61874 0.009655 open("/opt/gauss/gaussdata/user1", O_RDWR|O_SYNC <unfinished ...> 61874 0.000268 <... open resumed> ) = 31 59888 0.000045 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_TEMP2_UNDO_11_0.bak", O_RDWR|O_SYNC) = 36 61871 0.000040 open("/opt/gauss/gaussdata/temp2_undo", O_RDWR|O_SYNC) = 37 59888 0.000064 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_SYSAUX_12_0.bak", O_RDWR|O_SYNC) = 36 61871 0.007274 open("/opt/gauss/gaussdata/sysaux", O_RDWR|O_SYNC <unfinished ...> 61871 0.000811 <... open resumed> ) = 37 61870 0.000046 open("/opt/gauss/gaussdata/log1", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 30 61870 0.000052 open("/opt/gauss/gaussdata/log2", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 30 61870 0.000030 open("/opt/gauss/gaussdata/log3", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 30 61870 0.000024 open("/opt/gauss/gaussdata/log4", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 30 61870 0.000046 open("/opt/gauss/gaussdata/log5", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 30 61870 0.000025 open("/opt/gauss/gaussdata/log6", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 30 59888 0.000006 open("/tmp/gaussdb_backup/roger_level0_01.bak/arch_12_0.bak", O_RDWR|O_SYNC <unfinished ...> 59888 0.000008 <... open resumed> ) = 30 61871 0.000037 open("/opt/gauss/gaussdata/archive_log/arch_0_12.arc", O_RDWR|O_CREAT|O_EXCL|O_TRUNC|O_SYNC, 0600) = 31 61871 0.000036 open("/opt/gauss/gaussdata/data/cntl1", O_RDWR|O_SYNC) = 32 61871 0.000143 open("/opt/gauss/gaussdata/data/cntl2", O_RDWR|O_SYNC) = 33 61871 0.000024 open("/opt/gauss/gaussdata/data/cntl3", O_RDWR|O_SYNC) = 34 61871 0.000007 open("/opt/gauss/gaussdata/system", O_RDWR|O_SYNC) = 30 61872 0.000008 open("/opt/gauss/gaussdata/temp1_01", O_RDWR|O_SYNC) = 31 61873 0.000023 open("/opt/gauss/gaussdata/temp1_02", O_RDWR|O_SYNC <unfinished ...> 61873 0.015932 <... open resumed> ) = 35 61874 0.000069 open("/opt/gauss/gaussdata/undo", O_RDWR|O_SYNC) = 36 61872 0.000145 open("/opt/gauss/gaussdata/user1", O_RDWR|O_SYNC <unfinished ...> 61872 0.000746 <... open resumed> ) = 31 61873 0.000295 open("/opt/gauss/gaussdata/user2", O_RDWR|O_SYNC <unfinished ...> 61873 0.000058 <... open resumed> ) = 35 61871 0.000154 open("/opt/gauss/gaussdata/user3", O_RDWR|O_SYNC <unfinished ...> 61871 0.000042 <... open resumed> ) = 30 61874 0.000047 open("/opt/gauss/gaussdata/user4", O_RDWR|O_SYNC) = 36 61872 0.000066 open("/opt/gauss/gaussdata/user5", O_RDWR|O_SYNC) = 31 61872 0.000082 open("/opt/gauss/gaussdata/temp2_01", O_RDWR|O_SYNC <unfinished ...> 61872 0.000040 <... open resumed> ) = 31 61873 0.000123 open("/opt/gauss/gaussdata/temp2_02", O_RDWR|O_SYNC <unfinished ...> 61871 0.000084 open("/opt/gauss/gaussdata/temp2_undo", O_RDWR|O_SYNC <unfinished ...> 61873 0.001027 <... open resumed> ) = 30 61871 0.000030 <... open resumed> ) = 35 61872 0.000126 open("/opt/gauss/gaussdata/sysaux", O_RDWR|O_SYNC <unfinished ...> 61872 0.000046 <... open resumed> ) = 31 61871 0.000003 open("/opt/gauss/gaussdata/log1", O_RDWR|O_SYNC) = 30 61872 0.000004 open("/opt/gauss/gaussdata/log2", O_RDWR|O_SYNC <unfinished ...> 61872 0.000036 <... open resumed> ) = 31 61873 0.000031 open("/opt/gauss/gaussdata/log3", O_RDWR|O_SYNC <unfinished ...> 61873 0.000079 <... open resumed> ) = 35 61874 0.000051 open("/opt/gauss/gaussdata/log4", O_RDWR|O_SYNC <unfinished ...> 61874 0.000038 <... open resumed> ) = 36 61874 0.000169 open("/opt/gauss/gaussdata/log5", O_RDWR|O_SYNC <unfinished ...> 61871 0.000025 open("/opt/gauss/gaussdata/log6", O_RDWR|O_SYNC <unfinished ...> 61874 0.000098 <... open resumed> ) = 36 61871 0.000010 <... open resumed> ) = 30 61870 0.000129 open("/opt/gauss/gaussdata/log1", O_RDWR|O_SYNC) = 30 59888 0.000075 open("/opt/gauss/gaussdata/data/cntl1", O_RDONLY) = 17 59888 0.000047 open("/opt/gauss/gaussdata/data/cntl1", O_RDWR|O_SYNC <unfinished ...> 59888 0.000026 <... open resumed> ) = 17 59888 0.002681 open("/opt/gauss/gaussdata/data/cntl2", O_RDWR|O_SYNC <unfinished ...> 59888 0.000420 <... open resumed> ) = 17 59888 0.000415 open("/opt/gauss/gaussdata/data/cntl3", O_RDWR|O_SYNC) = 17 59888 0.000441 open("/opt/gauss/gaussdata/data/cntl2", O_RDWR|O_SYNC <unfinished ...> 59888 0.000207 <... open resumed> ) = 17 59888 0.000470 open("/opt/gauss/gaussdata/data/cntl3", O_RDWR|O_SYNC <unfinished ...> 59888 0.000189 <... open resumed> ) = 17 59888 0.000445 open("/opt/gauss/gaussdata/log1", O_RDWR|O_SYNC <unfinished ...> 59888 0.000224 <... open resumed> ) = 17 59888 0.000235 open("/opt/gauss/gaussdata/log2", O_RDWR|O_SYNC) = 18 59888 0.000430 open("/opt/gauss/gaussdata/log3", O_RDWR|O_SYNC) = 19 59888 0.000428 open("/opt/gauss/gaussdata/log4", O_RDWR|O_SYNC) = 20 59888 0.000466 open("/opt/gauss/gaussdata/log5", O_RDWR|O_SYNC) = 21 59888 0.000232 open("/opt/gauss/gaussdata/log6", O_RDWR|O_SYNC) = 22 59888 0.000046 open("/opt/gauss/gaussdata/system", O_RDWR|O_SYNC) = 23 59888 0.000370 open("/opt/gauss/gaussdata/data/cntl1", O_RDWR|O_SYNC) = 24 59888 0.000818 open("/opt/gauss/gaussdata/data/cntl2", O_RDWR|O_SYNC) = 25 59888 0.000243 open("/opt/gauss/gaussdata/data/cntl3", O_RDWR|O_SYNC) = 26 59888 0.000407 open("/opt/gauss/gaussdata/temp1_01", O_RDWR) = 27 59888 0.000036 open("/opt/gauss/gaussdata/temp1_02", O_RDWR) = 28 59888 0.001069 open("/opt/gauss/gaussdata/undo", O_RDWR|O_SYNC) = 29 59888 0.000492 open("/opt/gauss/gaussdata/user1", O_RDWR|O_SYNC) = 30 59888 0.000604 open("/opt/gauss/gaussdata/user2", O_RDWR|O_SYNC) = 31 59888 0.001332 open("/opt/gauss/gaussdata/user3", O_RDWR|O_SYNC <unfinished ...> 59888 0.000041 <... open resumed> ) = 32 59888 0.000023 open("/opt/gauss/gaussdata/user4", O_RDWR|O_SYNC) = 33 59888 0.000036 open("/opt/gauss/gaussdata/user5", O_RDWR|O_SYNC) = 34 59888 0.000433 open("/opt/gauss/gaussdata/temp2_01", O_RDWR|O_SYNC) = 35 59888 0.000240 open("/opt/gauss/gaussdata/temp2_02", O_RDWR|O_SYNC) = 36 59888 0.000039 open("/opt/gauss/gaussdata/temp2_undo", O_RDWR|O_SYNC) = 37 59888 0.000532 open("/opt/gauss/gaussdata/sysaux", O_RDWR|O_SYNC <unfinished ...> 59888 0.000027 <... open resumed> ) = 38 59888 0.000035 open("/opt/gauss/gaussdata/archive_log/arch_0_12.arc", O_RDWR|O_SYNC <unfinished ...> 59888 0.000022 <... open resumed> ) = 39 59888 0.000027 open("/opt/gauss/gaussdata/log1", O_RDWR|O_SYNC) = 40 59888 0.000351 open("/opt/gauss/gaussdata/log1", O_RDWR|O_SYNC) = 39 59888 0.000062 recvfrom(15, "\0\0\1\0\0\0\0\0\23\0\0\0alter database openo", 32, 0, NULL, NULL) = 32 59888 0.000294 open("/opt/gauss/gaussdata/archive_log/arch_0_12.arc", O_RDWR|O_SYNC) = 39 59888 0.000026 open("/opt/gauss/gaussdata/log1", O_RDWR|O_SYNC) = 40 59653 0.000017 open("/opt/gauss/gaussdata/system", O_RDWR|O_SYNC <unfinished ...> 59653 0.000009 <... open resumed> ) = 39 59662 0.000143 open("/opt/gauss/gaussdata/system", O_RDWR|O_SYNC) = 40 62916 0.000106 open("/opt/gauss/gaussdata/system", O_RDWR|O_SYNC <unfinished ...> 62916 0.000294 <... open resumed> ) = 41 62916 0.002062 open("/opt/gauss/gaussdata/undo", O_RDWR|O_SYNC <unfinished ...> 62915 0.000409 open("/opt/gauss/gaussdata/undo", O_RDWR|O_SYNC) = 42 62916 0.000762 <... open resumed> ) = 43 62915 0.000365 open("/opt/gauss/gaussdata/system", O_RDWR|O_SYNC) = 44 62915 0.000053 open("/opt/gauss/gaussdata/sysaux", O_RDWR|O_SYNC) = 45 62916 0.000694 open("/opt/gauss/gaussdata/sysaux", O_RDWR|O_SYNC <unfinished ...> 62916 0.000835 <... open resumed> ) = 46 62916 0.000038 open("/sys/devices/system/cpu/cpu0/topology/thread_siblings", O_RDONLY) = 47 62916 0.000040 open("/sys/devices/system/cpu/cpu0/topology/core_siblings", O_RDONLY) = 47 62916 0.000033 open("/proc/stat", O_RDONLY) = 47 62916 0.000109 open("/proc/vmstat", O_RDONLY) = 47 62916 0.000028 open("/proc/meminfo", O_RDONLY) = 47 62916 0.000035 open("/proc/loadavg", O_RDONLY) = 47 62916 0.009910 openat(AT_FDCWD, "/opt/gauss/gaussdata/log/longsql", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC <unfinished ...> 62916 0.000243 <... openat resumed> ) = -1 ENOENT (No such file or directory) [root@mysqldb tmp]# [root@mysqldb tmp]# [root@mysqldb tmp]# cat /tmp/gauss_recover.log |grep 'roger_level1_01' 
 

我们发现这里根本没有去读过增量备份集。这是为什么呢 ？ 那么只有一种解释；从日志来看。
 

gaussdb这里默认从0级别备份中去恢复了控制文件；而该控制文件肯定没有注册最新的增量信息，因此导致后面的增量无法应用。
 

由此可见应该是gaussdb在restore时默认是使用0级备份去进行恢复数据文件和控制文件；而不会寻找最新的控制文件备份。
 测试了几次，strace跟踪发现根本就不会读最新的备份集合；应该是控制文件惹的祸。 0级备份的控制文件中信息是比较旧的。按理说不该应该出现这种情况，可能是测试不对。改天再测一下。
 

备注：
 

若数据库可以恢复到CONSISTENT POINT，则可以保证数据一致性，ALTER DATABASE OPEN RESETLOGS可成功执行；
 否则，数据库已经不能保证数据一致性了，此时只能ALTER DATABASE OPEN FORCE IGNORE LOGS强行启动。
 

+++++ ==那么正确的恢复方法应该是什么呢 ？ ==+++++++++++++
 

[roger@mysqldb gaussdata]$ rm -rf user*
 [roger@mysqldb gaussdata]$ rm -rf temp*
 [roger@mysqldb gaussdata]$ rm -rf sys*
 [roger@mysqldb gaussdata]$ rm -rf undo*
 [roger@mysqldb gaussdata]$ rm -rf log1 log2 log3 log4 log5 log6
 [roger@mysqldb gaussdata]$ rm -rf data/cntl*
 [roger@mysqldb gaussdata]$ rm -rf archive_log/*
 

 

 

 rm -rf /gdb/gaussdb/data/data/*                
  rm -rf /gdb/gaussdb/data/archive_log/*                
 

 

 


 [roger@mysqldb gaussdata]$ ls -ltr
 total 4
 drwx------ 2 roger roger 42 Feb 13 16:23 dbs
 -rw------- 1 roger roger 0 Feb 13 16:23 zengine.lck
 drwx------ 2 roger roger 4096 Feb 13 21:50 trc
 drwx------ 5 roger roger 64 Feb 14 18:58 log
 drwx------ 2 roger roger 62 Feb 15 11:11 cfg
 drwx------ 2 roger roger 85 Feb 15 20:10 protect
 drwx------ 2 roger roger 6 Feb 16 13:23 data
 drwx------ 2 roger roger 6 Feb 16 13:23 archive_log
 

–restore（直接指定增量备份备份集；而不是指定全备的备份集）
 [roger@mysqldb gaussdata]$ zengine nomount -D /opt/gauss/gaussdata &
 [1] 127251
 [roger@mysqldb gaussdata]$ starting instance(nomount)
 instance started
 

[roger@mysqldb gaussdata]$
 [roger@mysqldb gaussdata]$ zsql / as sysdba -q
 

connected.
 

SQL> restore database from ‘/tmp/gaussdb_backup/roger_level1_01.bak’;
 

Succeed.
 

SQL> recover database;
 

Succeed.
 

SQL> alter database open;
 

Succeed.
 

SQL> select * from roger.test;
 

A B
 

 

1 www.enmotech.com
 1 www.killdb.com
 100 www.modb.pro
 

3 rows fetched.
 

可以看到可以恢复出最新的数据。我们可以看到restore时直接使用增量备集时可以恢复到最新时间点的数据的。
 

那么这种情况下，gaussdb会自动去从0级全备中先恢复文件；然后再应用增量吗 ？我们先来看下此时的日志：
 UTC+8 2020-02-16 13:28:53.494|ZENGINE|00000|127251|INFO>instance started
 UTC+8 2020-02-16 13:28:53.496|ZENGINE|00000|127266|INFO>job master thread started [srv_job.c:638]
 UTC+8 2020-02-16 13:29:50.222|ZENGINE|00000|128028|INFO>agent thread started, tid:139651462735616, close:0 [srv_agent.c:359]
 UTC+8 2020-02-16 13:29:50.222|ZENGINE|00053|128028|INFO>[LOG] file ‘/opt/gauss/gaussdata/log/audit/zengine.aud’ is added [cm_log.c:641]
 UTC+8 2020-02-16 13:29:56.323|ZENGINE|00053|128028|INFO>[BACKUP] restore start, device:0, policy:, paral count 4, path:/tmp/gaussdb_backup/roger_level1_01.bak [knl_restore.c:1947]
 UTC+8 2020-02-16 13:29:56.323|ZENGINE|00053|128028|INFO>[ARCH] Initialization complete [knl_archive.c:824]
 UTC+8 2020-02-16 13:29:56.331|ZENGINE|00000|128107|INFO>[BACKUP] parallel process 1 start [knl_bak_paral.c:996]
 UTC+8 2020-02-16 13:29:56.331|ZENGINE|00000|128108|INFO>[BACKUP] parallel process 2 start [knl_bak_paral.c:996]
 UTC+8 2020-02-16 13:29:56.333|ZENGINE|00000|128109|INFO>[BACKUP] parallel process 3 start [knl_bak_paral.c:996]
 UTC+8 2020-02-16 13:29:56.333|ZENGINE|00000|128110|INFO>[BACKUP] parallel process 4 start [knl_bak_paral.c:996]
 UTC+8 2020-02-16 13:29:56.409|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_0_0, recent filesize 0K, section [0K, 16528K], bakcup id 1 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:56.509|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_1_0, recent filesize 0K, section [0K, 16K], bakcup id 2 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:56.566|ZENGINE|00000|139650861757547|INFO>[RESTORE] fill datafile tail, start 16472K, end 16528K, backup id 1 [knl_bak_paral.c:763]
 UTC+8 2020-02-16 13:29:56.609|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_3_0, recent filesize 0K, section [0K, 393216K], bakcup id 3 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:56.611|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_3_1, recent filesize 0K, section [393216K, 531920K], bakcup id 4 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:56.830|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_4_0, recent filesize 0K, section [0K, 204240K], bakcup id 5 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:57.155|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_9_0, recent filesize 0K, section [0K, 16K], bakcup id 6 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:57.472|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_11_0, recent filesize 0K, section [0K, 16K], bakcup id 7 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:57.803|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_12_0, recent filesize 0K, section [0K, 89648K], bakcup id 8 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:58.836|ZENGINE|00000|139650861757549|INFO>[RESTORE] fill datafile tail, start 204192K, end 204240K, backup id 5 [knl_bak_paral.c:763]
 UTC+8 2020-02-16 13:29:59.549|ZENGINE|00000|128106|INFO>[RESTORE] restore datafiles finished [knl_restore.c:949]
 UTC+8 2020-02-16 13:29:59.635|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_0_0, recent filesize 16528K, section [0K, 16528K], bakcup id 1 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:59.736|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_1_0, recent filesize 16K, section [0K, 16K], bakcup id 2 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:59.736|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_3_0, recent filesize 531920K, section [0K, 393216K], bakcup id 3 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:59.837|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_3_1, recent filesize 531920K, section [393216K, 531920K], bakcup id 4 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:59.837|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_4_0, recent filesize 204240K, section [0K, 204240K], bakcup id 5 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:59.938|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_9_0, recent filesize 16K, section [0K, 16K], bakcup id 6 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:29:59.938|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_11_0, recent filesize 16K, section [0K, 16K], bakcup id 7 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:30:00.039|ZENGINE|00053|94592359855132|INFO>[RESTORE] restore data_12_0, recent filesize 89648K, section [0K, 90032K], bakcup id 8 [knl_bak_paral.c:374]
 UTC+8 2020-02-16 13:30:00.062|ZENGINE|00000|139650861757548|INFO>[RESTORE] fill datafile tail, start 89984K, end 90032K, backup id 8 [knl_bak_paral.c:763]
 UTC+8 2020-02-16 13:30:00.140|ZENGINE|00000|128106|INFO>[RESTORE] restore datafiles finished [knl_restore.c:949]
 UTC+8 2020-02-16 13:30:00.146|ZENGINE|00000|128106|INFO>[BACKUP] create logfiles completed [knl_restore.c:565]
 UTC+8 2020-02-16 13:30:00.157|ZENGINE|00000|139650861757547|INFO>[RESTORE] start restore arch_13_0, backup id 9, log name /tmp/gaussdb_backup/roger_level1_01.bak/arch_13_0.bak [knl_bak_paral.c:673]
 UTC+8 2020-02-16 13:30:00.160|ZENGINE|00000|139650861757547|INFO>[BACKUP] Create /opt/gauss/gaussdata/archive_log/arch_0_13.arc [knl_restore.c:362]
 UTC+8 2020-02-16 13:30:00.172|ZENGINE|00000|139650861757547|INFO>[ARCH] Record archive log file /opt/gauss/gaussdata/archive_log/arch_0_13.arc for log [0-13] start 0 end 1 [knl_archive.c:284]
 UTC+8 2020-02-16 13:30:00.250|ZENGINE|00000|128106|INFO>[RESTORE] restore logfiles finished, current backupset /tmp/gaussdb_backup/roger_level1_01.bak [knl_restore.c:964]
 UTC+8 2020-02-16 13:30:25.853|ZENGINE|00000|128107|INFO>[BACKUP] parallel process 1 stop [knl_bak_paral.c:1022]
 UTC+8 2020-02-16 13:30:25.855|ZENGINE|00000|128108|INFO>[BACKUP] parallel process 2 stop [knl_bak_paral.c:1022]
 UTC+8 2020-02-16 13:30:25.864|ZENGINE|00000|128109|INFO>[BACKUP] parallel process 3 stop [knl_bak_paral.c:1022]
 UTC+8 2020-02-16 13:30:25.875|ZENGINE|00000|128110|INFO>[BACKUP] parallel process 4 stop [knl_bak_paral.c:1022]
 UTC+8 2020-02-16 13:30:25.877|ZENGINE|00053|4995142158903276572|INFO>[BACKUP] restore success [knl_bak_common.c:1762]
 UTC+8 2020-02-16 13:30:43.488|ZENGINE|00053|206158558236|INFO>start to alter database MOUNT [knl_database.c:1047]
 UTC+8 2020-02-16 13:30:43.522|ZENGINE|00053|128028|INFO>[ARCH] Already initialized [knl_archive.c:803]
 UTC+8 2020-02-16 13:30:43.524|ZENGINE|00053|128028|INFO>sucessfully alter database MOUNT [knl_database.c:1103]
 UTC+8 2020-02-16 13:30:43.557|ZENGINE|00053|128028|INFO>database start recovery [knl_recovery.c:1457]
 UTC+8 2020-02-16 13:30:43.557|ZENGINE|00053|128028|INFO>recovery from file:13,point:750,lfn:27247 [knl_recovery.c:1459]
 UTC+8 2020-02-16 13:30:43.557|ZENGINE|00053|128028|INFO>recovery expected least end with file:13,point:750,lfn:27247 [knl_recovery.c:1461]
 UTC+8 2020-02-16 13:30:43.593|ZENGINE|00053|139650861757468|INFO>[RCY] recovery real end with file:14,point:1,lfn:27247 [knl_recovery.c:1423]
 UTC+8 2020-02-16 13:30:43.595|ZENGINE|00053|139650861757468|INFO>[RCY] current lfn 27247, rcy point lfn 27247, consistent point 27247, lrp point lfn 27247 [knl_recovery.c:1425]
 UTC+8 2020-02-16 13:31:40.820|ZENGINE|00053|206158558236|INFO>start to alter database OPEN [knl_database.c:1322]
 UTC+8 2020-02-16 13:31:40.944|ZENGINE|00053|139650861757468|INFO>[ARCH] Start ARCH thread for ARCHIVE_DEST_1[/opt/gauss/gaussdata/archive_log] [knl_archive.c:881]
 UTC+8 2020-02-16 13:31:40.946|ZENGINE|00053|2581544450367026204|INFO>The last shutdown is inconsistent [knl_database.c:1379]
 UTC+8 2020-02-16 13:31:40.946|ZENGINE|00053|128028|INFO>database start recovery [knl_recovery.c:1457]
 UTC+8 2020-02-16 13:31:40.946|ZENGINE|00053|128028|INFO>recovery from file:13,point:750,lfn:27247 [knl_recovery.c:1459]
 UTC+8 2020-02-16 13:31:40.946|ZENGINE|00053|128028|INFO>recovery expected least end with file:13,point:750,lfn:27247 [knl_recovery.c:1461]
 UTC+8 2020-02-16 13:31:40.957|ZENGINE|00053|139650861757468|INFO>[RCY] recovery real end with file:14,point:1,lfn:27247 [knl_recovery.c:1423]
 UTC+8 2020-02-16 13:31:40.957|ZENGINE|00053|139650861757468|INFO>[RCY] current lfn 27247, rcy point lfn 27247, consistent point 27247, lrp point lfn 27247 [knl_recovery.c:1425]
 UTC+8 2020-02-16 13:31:41.061|ZENGINE|00053|139650861757468|INFO>no valid standby configuration [knl_lns.c:1727]
 UTC+8 2020-02-16 13:31:41.066|ZENGINE|00053|128028|INFO>[DB] sse42 available 1 [knl_database.c:1440]
 UTC+8 2020-02-16 13:31:41.066|ZENGINE|00053|128028|INFO>sucessfully alter database OPEN [knl_database.c:1441]
 UTC+8 2020-02-16 13:31:41.125|ZENGINE|00000|127260|INFO>rollback thread closed [knl_tran.c:1203]
 UTC+8 2020-02-16 13:31:41.125|ZENGINE|00000|127261|INFO>rollback thread closed [knl_tran.c:1203]
 UTC+8 2020-02-16 13:31:41.653|ZENGINE|00054|129476|INFO>load index parts: load index parts, uid: 0, tid: 278, iid: 0, index partcnt: 4032 [knl_dc_part.c:993]
 UTC+8 2020-02-16 13:31:41.658|ZENGINE|00054|129476|INFO>load index parts: load index parts, uid: 0, tid: 302, iid: 0, index partcnt: 4032 [knl_dc_part.c:993]
 

可以看到日志中并没有记录有过去访问0级备份的信息。还好我再restore时开启了strace跟踪；下面看trace log：
 

[root@mysqldb bin]# strace -fr -F -o /tmp/strace_rec.log -p 127251
 Process 127251 attached with 16 threads
 Process 128028 attached
 Process 128106 attached
 Process 128107 attached
 Process 128108 attached
 Process 128109 attached
 Process 128110 attached
 Process 129475 attached
 Process 129476 attached
 Process 129477 attached
 Process 129478 attached
 Process 129479 attached
 ^CProcess 127251 detached
 Process 127252 detached
 Process 127253 detached
 Process 127254 detached
 Process 127255 detached
 Process 127256 detached
 Process 127257 detached
 Process 127258 detached
 Process 127259 detached
 Process 127262 detached
 Process 127263 detached
 Process 127264 detached
 Process 127265 detached
 Process 127266 detached
 Process 128028 detached
 Process 129475 detached
 

[root@mysqldb tmp]# cat strace_rec.log |grep ‘roger_level0’
 128028 0.000045 open("/tmp/gaussdb_backup/roger_level0_01.bak/backupset", O_RDWR|O_SYNC <unfinished …>
 128028 0.000023 open("/tmp/gaussdb_backup/roger_level0_01.bak/ctrl_0_0.bak", O_RDWR|O_SYNC) = 17
 128028 0.000004 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_SYSTEM_0_0.bak", O_RDWR|O_SYNC) = 30
 128028 0.000018 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_TEMP_1_0.bak", O_RDWR|O_SYNC) = 32
 128028 0.000021 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_UNDO_3_0.bak", O_RDWR|O_SYNC) = 30
 128028 0.000029 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_UNDO_3_1.bak", O_RDWR|O_SYNC <unfinished …>
 128028 0.000005 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_USERS_4_0.bak", O_RDWR|O_SYNC <unfinished …>
 128028 0.000005 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_TEMP2_9_0.bak", O_RDWR|O_SYNC <unfinished …>
 128028 0.000007 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_TEMP2_UNDO_11_0.bak", O_RDWR|O_SYNC) = 36
 128028 0.000004 open("/tmp/gaussdb_backup/roger_level0_01.bak/data_SYSAUX_12_0.bak", O_RDWR|O_SYNC <unfinished …>
 [root@mysqldb tmp]#
 

不难看出；确实会去寻找0级备份；然后再应用增量。这也验证了前面的分析。
 


                

        

        

    




    

      

        

            

              
                
                  jnrjian
                
              
            

            

                关注
              
            

        

        

          
    
            
  • 
              
                
                
                
                
                      0
                
              
              
    点赞
    
            
    • 
            
  • 
              
                
                
                
              
              
    踩
    
            
    • 
            
  • 
              
                
                
                
                
                    0
                
              
              
    
                
    
                  
                    收藏
                  
                
    
              
    
              
    
                
    
                  觉得还不错?
                  
                    一键收藏
                  
                 
                
    
              
    
            
    • 
          
  • 
            
    
              
              
            
    
              
              
              
                    0
              
            
            
    评论
    
          
    • 
          
  • 
          
    • 
          
  • 
            
              
            
              
    
            
              
                
            
    
          
    • 
        

      

      










                



	

		

			

				
					
Gaussdb(DWS)表备份到OBS，并可通过外表访问

				
			

			

				

					dba笔记
				

				

					05-06
					
					472
					
				

			

		

		

			
				
使用华为云OBS存储服务，将一些历史表备份至OBS，然后在数仓创建外表直接访问obs上的备份，这样就可以删除存在DWS上的历史表，释放空间，提高有效存储空间的利用率。

			
		

	



                

              
                



	

		

			

				
					
gaussdb 物理备份恢复【gs_rman 备份】【03】

				
			

			

				

					qq_42226855的博客
				

				

					08-26
					
					312
					
				

			

		

		

			
				
备份
gs_rman支持在线备份，它主要包含三种备份类型：全量备份、增量备份和备份WAL归档日志。在每一次备份后都应使用validate命令进行验证。
前提条件
已参照初始化配置完成备份目录初始化；在每个时间线上的增量备份和WAL归档日志备份之前必须至少完成了一次全量备份；gs_rman与gaussdb在部署同一台服务器上。

参数说明
备份时特有的参数

    
        
            参数名/缩写
            描述
        
    
    
	-b / --b

			
		

	



                


  
              

                


	

		

			

				
					
【GaussDB】普通备份恢复

				
			

			

				

					weixin_43346403的博客
				

				

					04-08
					
					338
					
				

			

		

		

			
				
5.启动数据库另外一种方法。3.恢复数据到磁盘。

			
		

	





	

		

			

				
					
GaussDB-物理、逻辑备份 使用方法和[GAUSS-53403]解决办法

				
			

			

				

					m0_46400195的博客
				

				

					11-25
					
					2439
					
				

			

		

		

			
				
数据库级备份包含数据库静态配置文件（cluster_static_config），数据库动态配置文件（cluster_dynamic_config），数据节点DN（Datanode）及其备实例。gs_dump是openGauss用于导出数据库相关信息的工具，用户可以自定义导出一个数据库或其中的对象（模式、表、视图等）。Roach也兼容使用python3 GaussRoach.py –t stop –F命令停止备份，有-F和无-F参数的执行结果相同。去全量备份的磁盘目录看下全量备份的名称后填写。

			
		

	



		

			
		



	

		

			

				
					
GaussDB灾备方案的设计

				
			

			

				

					Forlogenの解忧杂货铺
				

				

					01-10
					
					2293
					
				

			

		

		

			
				
文章目录1. 主数据中心/灾备数据中心2. 双运营数据中心3. 双活数据中心4. GaussDB同城双集群高可用方案5. 跨AZ容灾方案6. 参考
1. 主数据中心/灾备数据中心
在这种工作模式下，用户所有的业务系统在主数据中心中运行，而灾备数据中心为业务系统提供冷备或热备。当主数据中心的应用出现故障时，可以将单个应用或者数据中心整体切换到灾备数据中心。

2. 双运营数据中心
在这种工作模式下，一部分应用在第一个数据中心运行，另一部分应用在第二个数据中心运行，同时两个数据中心实现应用的互备，当某个应用出现

			
		

	





	

		

			

				
					
Rman增量备份Level012脚本

				
			

			

				

					03-12
				

			

		

		

			
				
Oracle Rman增量备份Level012脚本，公司用的oracle rac，asm的备份恢复都是通过RMAN，已实践！！

			
		

	





	

		

			

				
					
Oracle 12c备份恢复-RMAN工具技术手册

				
			

			

				

					12-27
				

			

		

		

			
				
Oracle 12c 备份恢复-RMAN 工具技术手册  本文档旨在提供 Oracle 12c 备份恢复的技术手册，主要介绍 RMAN 工具的概念、架构、备份类型和使用方法。  一、RMAN 概念  RMAN（Recovery Manager）是 Oracle 推荐的备份和...

			
		

	





	

		

			

				
					
顶级DBA漫谈Oracle Rman备份与恢复

				
			

			

				

					09-01
				

			

		

		

			
				
Oracle RMAN 备份与恢复概述  Oracle RMAN（Recovery Manager）是一种强大的备份和恢复工具，旨在帮助DBA管理员更好地管理和保护Oracle数据库。以下是Oracle RMAN备份与恢复的相关知识点：  备份恢复概述  备份恢复...

			
		

	





	

		

			

				
					
rman备份脚本和rman增量备份脚本分享

				
			

			

				

					09-10
				

			

		

		

			
				
主要介绍了rman备份脚本和rman增量备份脚本,需要的朋友可以参考下

			
		

	





	

		

			

				
					
rman备份策略

				
			

			

				

					07-13
				

			

		

		

			
				
自己用过的备份策略，简单实用，包含了rman0级备份，1级和2级备份的脚本。rman0级备份  #!/bin/bash  #chkinst.sh export EDITOR=vi export ORACLE_BASE=/u01/app/oracle ...

			
		

	





	

		

			

				
					
rman 1级备份

				
			

			

				

					cpgl40148的博客
				

				

					11-07
					
					419
					
				

			

		

		

			
				
1级增量备份是指自从上一次0级备份以来变化部分数据的备份。1级增量备份与0级备份共同完成
一个数据库恢复。注意，1级增量备份不能与全库备份一起完成一个数据库恢复。全库备份只能恢复到
自备份时间点数据库的状态。
[or...

			
		

	





	

		

			

				
					
GaussDB数据库的备份与恢复

					
最新发布

				
			

			

				

					Gauss松鼠会
				

				

					05-27
					
					1118
					
				

			

		

		

			
				
gs_dump是一款用于导出数据库相关信息的工具，支持导出完整一致的数据库对象（数据库、模式、表、视图等）数据，同时不影响用户对数据库的正常访问。

			
		

	





	

		

			

				
					
达梦 disql / as sysdba登录

				
			

			

				

					weixin_42018955的博客
				

				

					10-09
					
					2137
					
				

			

		

		

			
				
达梦 disql / as sysdba登录

			
		

	





	

		

			

				
					
Redhat 7.6安装GaussDB_100_1.0.1详细攻略

				
			

			

				

					weixin_41365682的博客
				

				

					05-22
					
					821
					
				

			

		

		

			
				
Redhat 7.6安装GaussDB_100_1.0.1详细步骤及白名单和开机自启动配置，数据库用户创建及远程登陆

			
		

	





	

		

			

				
					
GaussDB数仓高可用容灾利器之逻辑备份

				
			

			

				

					jnrjian的博客
				

				

					05-10
					
					386
					
				

			

		

		

			
				
技术解读丨GaussDB数仓高可用容灾利器之逻辑备份

华为云开发者社区

​

已认证帐号

摘要：GaussDB数仓的Roach工具，同时提供物理备份和逻辑备份两种主要形态的备份。逻辑备份针对数据库的逻辑对象进行抽取和备份，能够有效地应对单表、schema级等较细粒度的备份，较为灵活和便利。

一、简介

在大数据时代，数据的完整和可靠性成为一个数仓最核心的能力之一。GaussDB数仓以其出众的分布式计算和存储能力广受用户青睐的同时，也特别着眼于数据备份容灾领域的创新和打磨。数据的可靠性可以说是数仓的“

			
		

	





	

		

			

				
					
gaussdb 物理备份恢复【gs_rman 查看、删除备份】【05】

				
			

			

				

					qq_42226855的博客
				

				

					08-31
					
					400
					
				

			

		

		

			
				
1. 查看备份文件信息

查看、删除备份时特有选项


 
 
 
  参数名
  描述
 
 
  timeline
  查看备份文件的时间线。
 
 
  DATE
  查看指定的备份文件的详细信息。
    取值范围：备份的开始时间，即使用gs_rman show命令查看到的开始时间（start）。
    默认值：如果没有指定，将持续还原到最新的备份时间。
 


查看、删除备份时公共参数

    
        
            参数名/缩写
            描述
      

			
		

	





	

		

			

				
					
gaussdb 逻辑备份恢复【gs_dump 简介、参数说明】【01】

				
			

			

				

					qq_42226855的博客
				

				

					09-01
					
					2036
					
				

			

		

		

			
				
gs_dump 简介
       gs_dump是GaussDB用于备份数据库相关信息的逻辑备份工具。它可以对正在运行的数据库进行完整的一致备份，支持将数据库中的数据定义和数据导出（转储）成物理文件，并使用导入工具进行恢复，同时还支持对数据库的状态信息及syslog进行导出。
三大功能：
1.导出格式可以是一个文本（可加密）或目录格式
2.用户可通过gs_dump连接到指定数据库，对数据库进行状态信息查询，并将查询结果显示出来
3.用户

			
		

	





	

		

			

				
					
oracle数据库备份报错RMAN returned exit code 1

				
			

			

				

					02-14
				

			

		

		

			
				
在 Oracle 数据库备份过程中，如果出现 RMAN 返回退出代码 1 的错误，通常意味着备份操作失败。原因可能有很多，例如：  1. 数据库不可用，导致无法连接到数据库。 2. 磁盘空间不足，导致无法存储备份文件。 3. RMAN...

			
		

	



              



  
“相关推荐”对你有帮助么？

  

      
    
          
  • 
              
    
                  
                  
              
    
              
    非常没帮助
    
          
    • 
          
  • 
              
    
                  
                  
              
    
              
    没帮助
    
          
    • 
          
  • 
              
    
                  
                  
              
    
              
    一般
    
          
    • 
          
  • 
              
    
                  
                  
              
    
              
    有帮助
    
          
    • 
          
  • 
              
    
                  
                  
              
    
              
    非常有帮助
    
          
    • 
      

      

      

          
          提交
      

      

  




          




        



    





    



      

      

        
      

      





	

		评论	
	

	
	




  

    

      添加红包
      
    

    

      

        
        

          
          
        

        
请填写红包祝福语或标题

      

      

        
        

          
          个
        

        
红包个数最小为10个

      

      

        
        

          
          元
        

        
红包金额最低5元

      

      

        
        

          当前余额3.43元
          前往充值 >
        

      

      

        

          需支付：10.00元
        

        
        
      

    

  





  

    
    
  

  

    
    
  








  

    

      

        

          

            
            
成就一亿技术人!

          

          

        

        

          

          

            领取后你会自动成为博主和红包主的粉丝
            规则
          

        

      

      

        

          

            
              
            
            

              
hope_wisdom
 发出的红包
            

          

        

        

          

          

          

        

      

    

    

  



      
      

      
实付元

      
使用余额支付

      

        

          

            
            点击重新获取
          

        

        
扫码支付

      

      

        
          
            
          
          
        
      

      

        
        钱包余额
          0
          

            
            

              

                
抵扣说明：

                
 1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
 2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

              

            

          

      

      余额充值