Question about hardware system information

[Itamar Heim]

On 07/15/2013 10:07 PM, Gustavo Frederico Temple Pedrosa wrote:
>> -----Original Message-----
>> From: Itamar Heim [mailto:iheim at redhat.com]
>> Sent: segunda-feira, 15 de julho de 2013 10:15
>> To: Gustavo Frederico Temple Pedrosa
>> Cc: arch at ovirt.org
>> Subject: Re: Question about hardware system information
>>
>> On 07/15/2013 03:55 PM, Gustavo Frederico Temple Pedrosa wrote:
>>> Hello,
>>>
>>> The VDSM uses Dmidecode to retrieve the information that will be used
>>> in the method
>> VdsBrokerObjectsBuilder.UpdateHardwareSystemInformation().
>>>
>>> The IBM POWER architecture does not have the information from the
>>> Dmidecode, however, it has other information.
>>>
>>> In the case of new information for IBM POWER architecture, there are
>>> two
>>> options: (1) add new columns in the method
>> VdsDAODbFacadeImpl.mapRow()
>>> and in the database, or (2) map the new set of information into a
>>> single column (using, for example, JSON).
>>>
>>> What would be the best option? And in case of a future new architecture?
>>>
>>> Regards,
>>>
>>> Gustavo F Temple Pedrosa.
>>>
>>>
>>>
>>> _______________________________________________
>>> Arch mailing list
>>> Arch at ovirt.org
>>> http://lists.ovirt.org/mailman/listinfo/arch
>>>
>>
>> can you give an example of how different are the fields? we take basic fields
>> iirc - vendor, serial number, etc.?
>
>
> - Examples:
>      System Id: IBM,0310050AA (/proc/device-tree/system-id)
>      Platform: pSeries (/proc/cpuinfo)
>      Model: IBM,8246-L2D (/proc/cpuinfo)
>      Machine: CHRP IBM,8246-L2D (/proc/cpuinfo)
>

ok - these seem to be identical to what we do with dmidecode, why can't 
you just send these in same way/fields and engine will just show them 
the same?