Re: Lot of Latch Free and Enqueue Hash Chain Wait Events

From: Stefan Knecht <knecht.stefan_at_gmail.com>
Date: Wed, 4 Oct 2017 01:13:11 +0700
Message-ID: <CAP50yQ89ydBvozMywoCe20B1ZjkZ6EDSse4TA-j7Q-uh=e7LJQ_at_mail.gmail.com>



Yes, exactly as Niall said.

The CPU limit is hard-coded into the software and they implemented it through the resource manager. Whenever you're actively using more than 16 threads, resource manager will prevent your from getting any more CPU until a thread has been freed.

I was bitten by this too when testing (deliberately producing) latch contention behavior. You can't really put too much concurrent stress on an SE2 instance, you'll simply be throttled.

Stefan

On Wed, Oct 4, 2017 at 12:07 AM, Niall Litchfield < niall.litchfield_at_gmail.com> wrote:

> SE2 is core limited by if memory serves CPU_COUNT.
>
> On 3 Oct 2017 17:09, "Mladen Gogala" <gogala.mladen_at_gmail.com> wrote:
>
>> Hi Stefan, can you expand a bit on this?
>>
>> Regards
>>
>>
>> On 10/03/2017 11:44 AM, Stefan Knecht wrote:
>>
>>> Is this Standard Edition 2 ? You may be throttled by the database on
>>> purpose.
>>>
>>
>> --
>> Mladen Gogala
>> Oracle DBA
>> Tel: (347) 321-1217
>>
>> --
>> http://www.freelists.org/webpage/oracle-l
>>
>>
>>

-- 
//
zztat - The Next-Gen Oracle Performance Monitoring and Reaction Framework!
Visit us at zztat.net | Support our Indiegogo campaign at igg.me/at/zztat |
_at_zztat_oracle

--
http://www.freelists.org/webpage/oracle-l
Received on Tue Oct 03 2017 - 20:13:11 CEST

Original text of this message