Re: intermittent long "log file sync" waits

Yep seems so, at least judging from the parameter names:

SQL> _at_pd adaptive%log
Show all parameters and session values from x$ksppi/x$ksppcv...

       NUM N_HEX NAME
VALUE

---------- ----- --------------------------------------------------------
------
      1885   75D *_adaptive_scalable_log_writer*_enable_worker_threshold
 200
      1886   75E *_adaptive_scalable_log_writer*_disable_worker_threshold
50
      1887   75F *_adaptive_scalable_log_writer*_sampling_count
128
      1888   760 *_adaptive_scalable_log_writer*_sampling_time
 3
      1889   761 *_adaptive_scalable_log_writer*_enable_worker_aging
 999900
      1915   77B _use_adaptive_log_file_sync
 TRUE
      1916   77C _adaptive_log_file_sync_use_polling_threshold
 110
      1917   77D _adaptive_log_file_sync_use_postwait_threshold
50
      1918   77E _adaptive_log_file_sync_use_postwait_threshold_aging
1001
      1919   77F _adaptive_log_file_sync_sampling_count
128
      1920   780 _adaptive_log_file_sync_sampling_time                    3

      1921   781 _adaptive_log_file_sync_sched_delay_window
60
      1922   782 _adaptive_log_file_sync_poll_aggressiveness              0

      1923   783 _adaptive_log_file_sync_high_switch_freq_threshold       3

--
Tanel Poder
https://tanelpoder.com/seminar


On Mon, Feb 3, 2020 at 4:38 AM Frits Hoogland <frits.hoogland_at_gmail.com>
wrote:



> I do believe the automatic switching between the logwriter doing


> everything and using worker processes is called ‘adaptive scalable

> logwriter’.

>

> Frits Hoogland

>

> http://fritshoogland.wordpress.com

> frits.hoogland_at_gmail.com

> Mobile: +31 6 14180860

>

> On 2 Feb 2020, at 23:09, Tanel Poder <tanel_at_tanelpoder.com> wrote:

>

> Btw, adaptive log file sync is a different thing than the adaptive switch

> between LGWR-only processing vs. LGWR+LG% slave processing (that's why I

> mentioned this parameter:  _use_single_log_writer=adaptive)

>

>

>


--
http://www.freelists.org/webpage/oracle-l