Oracle V$SYSTEM_EVENT字段等解释

Oracle V$SYSTEM_EVENT

V$SYSTEM_EVENT

V$SYSTEM_EVENT displaysinformation on total waits for an event. Note thatthe TIME_WAITED and AVERAGE_WAIT columnswill contain a value of zero on those platforms that do not supporta fast timing mechanism. If you are running on one of theseplatforms and you want this column to reflect true wait times, thenyou must set TIMED_STATISTICS to TRUE inthe parameter file; doing this will have a small negative effect onsystem performance.

See Also:

"TIMED_STATISTICS"

Column	Datatype	Description
EVENT	VARCHAR2(64)	Name of the wait event
TOTAL_WAITS	NUMBER	Total number of waits for the event
TOTAL_TIMEOUTS	NUMBER	Total number of timeouts for the event
TIME_WAITED	NUMBER	Total amount of time waited for the event(in hundredths of a second)
AVERAGE_WAIT	NUMBER	Average amount of time waited for the event(in hundredths of a second)
TIME_WAITED_MICRO	NUMBER	Total amount of time waited for the event(in microseconds)
TOTAL_WAITS_FG	NUMBER	Total number of waits for the event, fromforeground sessions
TOTAL_TIMEOUTS_FG	NUMBER	Total number of timeouts for the event, fromforeground sessions
TIME_WAITED_FG	NUMBER	Amount of time waited for the event (inhundredths of a second), from foreground sessions
AVERAGE_WAIT_FG	NUMBER	Average amount of time waited for the event(in hundredths of a second), from foreground sessions
TIME_WAITED_MICRO_FG	NUMBER	Amount of time waited for the event (inmicroseconds), from foreground sessions
EVENT_ID	NUMBER	Identifier of the wait event
WAIT_CLASS_ID	NUMBER	Identifier of the class of the waitevent
WAIT_CLASS#	NUMBER	Number of the class of the wait event
WAIT_CLASS	VARCHAR2(64)	Name of the class of the wait event

本视图概括了实例各项事件的等待信息。v$session_wait显示了系统的当前等待项，v$system_event则提供了自实例启动后各个等待事件的概括。常用于获取系统等待信息的历史影象。而通过两个snapshot获取等待项增量，则可以确定这段时间内系统的等待项。

 

V$SYSTEM_EVENT中的常用列

l        EVENT:等待事件名称

l        TOTAL_WAITS：此项事件总等待次数

l        TIME_WAITED：此项事件的总等待时间(单位：百分之一秒)

l        AVERAGE_WAIT：此项事件的平均等待用时(单位：百分之一秒)(time_waited/total_waits)

l        TOTAL_TIMEOUTS：此项事情总等待超时次数

注释：

存在等待（某事），就可能会产生等待的该事没来，即所谓等待超时。

 

示例：

1.查看系统的各项等待，按总耗时排序

SELECT event,total_waits waits,total_timeouts timeouts,

      time_waited total_time,average_wait avg

FROM V$SYSTEM_EVENT

ORDER BY 4 DESC;

 

比如，通过checkpoint completed、log file switch（checkpointincomplete）可以查看检查点进程的性能。通过log file parallel write、log file switchcompleted可以查看联机重做日志文件的性能。通过log file switch（archivingneeded）事件可以检查归档进程的性能。

 

找出瓶颈：

1。通过Statspack列出空闲事件。

2。检查不同事件的等待时间开销。

3。检查每条等待记录的平均用时，因为某些等待事件(比较log file switchcompletion)可能周期性地发生，但发生时却造成了严重的性能损耗。