使用eventapi診斷數據導入imp性能

　　SQL> select eventTOTAL_WAITS TIME_WAITEDAVERAGE_WAIT from v$session_event where sid= order by TIME_WAITED desc;

　　EVENT                                                            TOTAL_WAITS

TIME_WAITED AVERAGE_WAIT

db file sequential read

　　free buffer waits

　　log file switch completion

　　EVENT                                                            TOTAL_WAITS

TIME_WAITED AVERAGE_WAIT

log file sync

　　SQL*Net message from client

　　latch free

　　EVENT                                                            TOTAL_WAITS

TIME_WAITED AVERAGE_WAIT

log file switch (checkpoint incomplete)

　　SQL*Net more data from client

　　enqueue

　　EVENT                                                            TOTAL_WAITS

TIME_WAITED AVERAGE_WAIT

undo segment extension

　　write complete waits

　　buffer busy waits

　　EVENT                                                            TOTAL_WAITS

TIME_WAITED AVERAGE_WAIT

SQL*Net message to client

　　file open

　　發現db file sequential read 居高通常在單塊讀發生該事件用於索引讀取察看正在導入數據的表果然索引俱全導入數據的時候要維護索引對每個導入的數據都要找到對應的索引葉結點插入新索引enry 刪除索引後該等待事件降低

　　通常如果發生較高的log file sync 事件表示導入進程提交過於頻繁增加buffer參數可以減少commit次數減少該等待事件

From:http://tw.wingwit.com/Article/program/Oracle/201311/18307.html