ngle write 2 0 0 2
rdbms ipc message 3,494 84 8,851 2533
DIAG idle wait 1,088 100 1,089 1001
smon timer 2 100 600 ######
MRP redo arrival 623 0 544 873
pmon timer 181 100 543 3002
shared server idle wait 18 100 540 30007
dispatcher timer 9 100 540 60005
-------------------------------------------------------------
Wait Event Histogram DB/Inst: dg2/dg2 Snaps: 1-2
-> Total Waits - units: K is 1000, M is 1000000, G is 1000000000
-> % of Waits - column heading: <=1s is truly <1024ms, >1s is truly >=1024ms
-> % of Waits - value: .0 indicates value was <.05%, null is truly 0
-> Ordered by Event (idle events last)
Total ----------------- % of Waits ------------------
Event Waits <1ms <2ms <4ms <8ms <16ms <32ms <=1s >1s
-------------------------- ----- ----- ----- ----- ----- ----- ----- ----- -----
Data file init write 24 75.0 16.7 4.2 4.2
Disk file operations I/O 1096 99.4 .2 .2 .3
RFS dispatch 622 99.8 .2
RFS ping 9 88.9 11.1
RFS random i/o 543 99.8 .2
RFS write 613 96.1 2.6 .8 .3 .2
SQL*Net more data to clien 6 100.0
Standby redo I/O 543 .2 27.8 37.0 5.9 3.1 13.3 12.7
asynch descriptor resize 1 100.0
control file parallel writ 587 47.0 26.7 4.6 3.2 11.1 7.3
control file sequential re 5657 100.0 .0
db file async I/O submit 24 4.2 12.5 8.3 45.8 29.2
db file parallel read 4 25.0 25.0 50.0
db file sequential read 31 58.1 16.1 9.7 16.1
db file single write 2 50.0 50.0
latch free 104 99.0 1.0
latch: row cache objects 1 100.0
latch: shared pool 1 100.0
library cache lock 8 12.5 50.0 12.5 25.0
log file sequential read 1087 100.0
os thread startup 1 100.0
DIAG idle wait 1088 100.0
MRP redo arrival 623 .3 .8 1.8 1.1 1.8 75.8 18.5
SQL*Net message from clien 895 7.2 12.8 8.3 2.6 3.1 1.2 53.9 10.9
SQL*Net message to client 895 100.0
SQL*Net more data from cli 11 63.6 9.1 9.1 9.1 9.1
class slave wait 1 100.0
dispatcher timer 9 100.0
pmon timer 181 100.0
rdbms ipc message 3493 .1 .1 .1 .3 .5 .5 63.4 35.1
shared server idle wait 18 100.0
smon timer 2 100.0
-------------------------------------------------------------
。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。。
那么备库的性能诊断，就有了利器。
另外，我们最希望的是自动去执行收集统计信息，那么statspack有spauto.sql去创建job，那么对于stdby也可以创建job使其自动去执行。
参考我的另一篇日志《oracle statspack学习》http://blog.csdn.net/rhys_oracle/article/details/11694355
过程如下：
SQL> show parameter job_queue_processes
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
job_queue_processes integer 1000
SQL> show parameter timed_statistics
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
timed_statistics boolean TRUE
SQL>
SQL> col owner for a14
SQL> col db_link for a20
SQL> col host for a18
SQL> col username for a40
SQL> select * from dba_db_links;
OWNER DB_LINK USERNAME
-------------- -------------------- ----------------------------------------
HOST CREATED
------------------ ---------
STDBYPERF STDBY_LINK_DG2 PERFSTAT
DG2 23-APR-14
SQL> set linesize 200
SQL> r
1* select * from dba_db_links
OWNER DB_LINK USERNAME HOST CREATED
-------------- -------------------- ---------------------------------------- ------------------ ---------
STDBYPERF STDBY_LINK_DG2 PERFSTAT DG2 23-APR-14
SQL> conn stdbyperf/Amy
Connected.
SQL> select instance_name from v$instance@STDBY_LINK_DG2;
INSTANCE_NAME
----------------
dg2
SQL>
创建job：
SQL> select job,schema_user,last_date,what,instance from u