Did the wait event show up, after some trace files were generated? I know I am not helping, but I have been there. My 9.2.0.6 RAC on Solaris 8 experienced the same symptoms . We ran into an ORA-4031 and some trace files were written into, and since then, the DIAG process was forever waiting on "async disk i/o". This did not cause any impact, other than skewing up our monitoring. A bounce of the instance fixed this. Had that luxury since this was QA. Sami Seerangan <dba.orcl@xxxxxxxxx> Sent by: oracle-l-bounce@xxxxxxxxxxxxx 09/14/2005 09:42 PM Please respond to dba.orcl To: Rajesh.Rao@xxxxxxxxxxxx cc: oracle-l <oracle-l@xxxxxxxxxxxxx> Subject: Re: async disk IO in RAC DIAG process is experiencing all these waits. On 9/14/05, Rajesh.Rao@xxxxxxxxxxxx < Rajesh.Rao@xxxxxxxxxxxx> wrote: What's the process experiencing all these waits? Is it DIAG or DBWR?