课程第15次（2018-11-13星期二）

botang

The fast-start checkpointing feature is enabled by specifying a non-zero desired mean-time to recover (MTTR) value, which will be used to set the FAST_START_MTTR_TARGET initialization parameter. This parameter controls the amount of time the database takes to perform crash recovery for a single instance. When fast-start checkpointing is enabled, Oracle automatically maintains the speed of checkpointing so that the requested MTTR is achieved. Setting the value to 0 will disable this functionality.

1. SQL> alter system set fast_start_mttr_target=10;
   
2. 
   
3. System altered.
   
4. 
   
5. SQL> show parameter log_check
   
6. 
   
7. NAME                                     TYPE         VALUE
   
8. ------------------------------------ ----------- ------------------------------
   
9. log_checkpoint_interval              integer         0
   
10. log_checkpoint_timeout                     integer         1800
    
11. log_checkpoints_to_alert             boolean         FALSE
    
12. SQL> alter system set log_checkpoints_to_alert=true;
    

复制代码

1.   Current log# 2 seq# 122 mem# 0: +DATA/orcl/onlinelog/group_2.262.816169639
   
2.   Current log# 2 seq# 122 mem# 1: +FRA/orcl/onlinelog/group_2.258.816169639
   
3. Tue Nov 13 18:56:22 2018
   
4. Archived Log entry 113 added for thread 1 sequence 121 ID 0x501b221c dest 1:
   
5. Tue Nov 13 21:06:49 2018
   
6. ALTER SYSTEM SET fast_start_mttr_target=3601 SCOPE=BOTH;
   
7. Tue Nov 13 21:10:45 2018
   
8. MTTR advisory was temporarily turned off because FAST_START_MTTR_TARGET was altered.
   
9. ALTER SYSTEM SET fast_start_mttr_target=10 SCOPE=BOTH;
   
10. Tue Nov 13 21:11:09 2018
    
11. FAST_START_MTTR_TARGET 10 is set too low, using minimum achievable MTTR 17 instead.
    
12. Tue Nov 13 21:12:28 2018
    
13. ALTER SYSTEM SET log_checkpoints_to_alert=TRUE SCOPE=BOTH;
    
14. 
    
15. 
    
16. 
    
17. 
    
18. Tue Nov 13 21:24:56 2018
    
19. MTTR advisory was temporarily turned off because FAST_START_MTTR_TARGET was altered.
    
20. ALTER SYSTEM SET fast_start_mttr_target=5 SCOPE=BOTH;
    
21. Tue Nov 13 21:25:15 2018
    
22. Incremental checkpoint up to RBA [0x7a.b1e3.0], current log tail at RBA [0x7a.b3d6.0]
    
23. FAST_START_MTTR_TARGET 5 is set too low, using minimum achievable MTTR 17 instead.
    

复制代码

当下一个Incremental checkpoint up to RBA ,
  select d.RESETLOGS_CHANGE#,
            d.CHECKPOINT_CHANGE#,
            d.CURRENT_SCN
    from v_$database  d;
   
    select  *
     from v_$datafile df;

checkpoint_change#才会更新。(这个号，用来表征对应的数据文件，在内存中再也找不到比它小的块)。




上完1Z0-052介绍章（12/40），1Z0-053介绍章（13/40）

jarlean

老师，单机测试环境（版本：11.0.2.1），实验没成功，Incremental checkpoint已经发生3次了，但是CHECKPOINT_CHANGE#还是没有变化。

实验过程：
1.记录发生Incremental checkpoint前的CHECKPOINT_CHANGE#信息。


2.查看trace日志，checkpoint发生


3.Incremental checkpoint发生后，CHECKPOINT_CHANGE#未发生变化


4.第二次Incremental checkpoint产生，但是CHECKPOINT_CHANGE#还是没有发生改变。


5.第三次Incremental checkpoint产生，但是CHECKPOINT_CHANGE#还是没有发生改变。

botang

jarlean 发表于 2018-11-14 11:36
老师，单机测试环境（版本：11.0.2.1），实验没成功，Incremental checkpoint已经发生3次了，但是CHECKPOIN ...

继续等待，下一次Incremental Checkpoint就会更新。

jarlean

botang 发表于 2018-11-15 07:58
继续等待，下一次Incremental Checkpoint就会更新。