Re: Strange problem

  • From: Mladen Gogala <gogala.mladen@xxxxxxxxx>
  • To: fairlie.rego@xxxxxxxxx
  • Date: Mon, 1 May 2017 00:34:02 -0400

On 04/30/2017 06:13 PM, fairlie.rego@xxxxxxxxx wrote:

Are you seeing this behaviour if you set

_datafile_write_errors_crash_instance = FALSE (not default) since this is what I do


Hi Fairlie,

I haven't tried that, but I will. Thanks for the advice. For now, I've set filesystemio_options to none, increased the number of db_writer_processes to 2 and dbwr_slaves to 4. What I did notice is the expected increased consumption of memory, since the blocks are now cached in the buffer cache. The nightly batch is still running, the night is still young.


Mladen Gogala
Oracle DBA
Tel: (347) 321-1217

Other related posts: