Re: Another licensing Q

From: Alfredo Abate <alfredo.abate_at_gmail.com>
Date: Mon, 6 Nov 2017 16:11:34 -0600
Message-ID: <CALrB5poqyX985-kErRh_EgW4UaongCrqmu4cVP4-XP_Zz4v3Dw_at_mail.gmail.com>



The document that Stefan has presented is what I was remembering from a while back when it comes to disabling cores in the BIOS.

In the past, I have physically pulled CPU sockets out of the server to get the server where it needed to be. Luckily server CPU utilization was low enough to begin with that it didn't impact the system once the sockets were removed. I wouldn't recommend this route unless you know if your application/database can sustain with a reduced horsepower.

Alfredo

On Mon, Nov 6, 2017 at 3:08 PM, Stefan Knecht <knecht.stefan_at_gmail.com> wrote:

> This is actually documented by Oracle, here:
>
> http://www.oracle.com/us/corporate/pricing/partitioning-070609.pdf
>
> In a nutshell, disabling something in a BIOS is not an approved method of
> reducing the number of cores. You need to use one of the listed methods and
> you'll be fine.
>
>
> Stefan
>
>
>
>
> On Tue, Nov 7, 2017 at 12:26 AM, Atkinson, Matthieu <
> matthieu.atkinson_at_gmail.com> wrote:
>
>> Hello,
>>
>> I have customers running OVM on non-Oracle hardware that have
>> successfully completed an Oracle audit... as a matter of fact, in france at
>> least, there's an increase demand in OVM deployments outside of ODA.
>>
>> On Mon, Nov 6, 2017 at 3:37 PM, Andrew Kerber <andrew.kerber_at_gmail.com>
>> wrote:
>>
>>> I am pretty sure that Oracle will only accept that method on Oracle
>>> hardware, ie ODA. You will probably need to license all of them on any
>>> other hardware. That is not any sort of official answer of course, but its
>>> my best guess.
>>>
>>> On Mon, Nov 6, 2017 at 2:41 AM, Marian Bednar <bednar_at_nbs.sk> wrote:
>>>
>>>> Hi Thomas,
>>>>
>>>> only note - on bare-metal servers output form “cat /proc/cpuinfo|grep
>>>> processor|wc –l” usually show the number of threads (if multithreading
>>>> is on) then number of cores is half of it. Try command "lscpu".
>>>> Number of cores you can also see in oracle alert.log during instance
>>>> startup (since 11.2.0.4), e.g.
>>>>
>>>> Initial number of CPU is 16
>>>> Number of processor cores in the system is 8
>>>> Number of processor sockets in the system is 2
>>>>
>>>> Marian
>>>>
>>>>
>>>>
>>>> From: Biju Thomas <biju.thomas_at_gmail.com>
>>>> To: "oracle-l_at_freelists.org" <oracle-l_at_freelists.org>,
>>>> Date: 03. 11. 2017 20:38
>>>> Subject: Another licensing Q
>>>> Sent by: oracle-l-bounce_at_freelists.org
>>>> ------------------------------
>>>>
>>>>
>>>>
>>>> Question related to Oracle CPU licensing on Cisco UCS. My current blade
>>>> has 32 cores (the result of “cat /proc/cpuinfo|grep processor|wc –l”).
>>>> The sysadmin disabled all but four cores on each socket via the UCS BIOS
>>>> and got the output of 8 when I ran the same command the second time. The OS
>>>> only sees 8. The question is does Oracle allow me to license only 8 visible
>>>> cores, or do I have to still license all 32 present on the blade?
>>>>
>>>> I believe such licensing is possible on ODA (Oracle Database
>>>> Appliance).
>>>> *https://docs.oracle.com/cd/E22693_01/doc.12/e25375/chapter1.htm*
>>>> <https://docs.oracle.com/cd/E22693_01/doc.12/e25375/chapter1.htm>
>>>> With bare-metal, you disable cores that will not be used by the Oracle
>>>> Database by adding your hardware Support Identifier (SI) for Oracle
>>>> Database Appliance to your My Oracle Support account and creating a key.
>>>>
>>>>
>>>> Thanks much!
>>>> Biju Thomas
>>>>
>>>>
>>>> --
>>>> Best,
>>>> Biju Thomas
>>>> *www.bijoos.com* <http://www.bijoos.com/>
>>>>
>>>
>>>
>>>
>>> --
>>> Andrew W. Kerber
>>>
>>> 'If at first you dont succeed, dont take up skydiving.'
>>>
>>
>>
>
>
> --
> //
> 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 Mon Nov 06 2017 - 23:11:34 CET

Original text of this message