This event occurs in the DBWR. It indicates that the DBWR is performing a parallel write to files and blocks. When the last I/O has gone to disk, the wait ends. This wait only occurs in the DBWR processes. Problem Ultimately, the operating system's I/O subsystem is the bottleneck. User sessions wait for DBWR to complete the process of flushing dirty blocks from the buffer cache. To further determine user sessions that are also impacted, look for long wait times on "write complete waits" and / or "free buffer waits". Solution Two possible solutions for extensive waits on db file parallel write are to Improve DBWR Throughput or to reduce I/O rate demands. This wait may also stem unnecessarily frequent flushing of the buffer cache by the DBWR. Have a look at the view V$INSTANCE_RECOVERY to see if the target recovery time is leading to excessive writing of blocks from the cache.
本文转自maclean_007 51CTO博客，原文链接:http://blog.51cto.com/maclean/1277907
