概述

一大早运维团队就来找事，说系统又有点卡了，然后发现了一个比较少见的等待事件–kfk: async disk IO，趁着这次排查的过程也简单说下这个等待事件吧!

1、查看TOP N等待事件

SELECTinst_id,EVENT,SUM(DECODE(WAIT_TIME,0,0,1))"Prev",SUM(DECODE(WAIT_TIME,0,1,0))"Curr",COUNT(*)"Tot",
sum(SECONDS_IN_WAIT)SECONDS_IN_WAIT
FROMGV$SESSION_WAIT
WHEREeventNOT
IN('smontimer','pmontimer','rdbmsipcmessage','SQL*Netmessagefromclient','gcsremotemessage')
ANDeventNOTLIKE'%idle%'
ANDeventNOTLIKE'%Idle%'
ANDeventNOTLIKE'%StreamsAQ%'
GROUPBYinst_id,EVENT
ORDERBY1,5desc;
--classslavewait