David Hollister wrote:
>>>> With S10U4+patches, PCIe SAS card, I get:
>>>>         
>>> SAS ?! Can this be a problem ? AFAIK, fcinfo looks
>>> specifically for FC HBA. (In fact it uses
>>>       
>> libHBAAPI,
>>     
>>> which routes the actual call via /etc/hba.conf)
>>> Odds are that your HBA does not provide a library
>>> for libHBAAPI.
>>>       
>
> That is correct.  I'm assuming Cyril has a 2540.  He is seeing SAS/SATA 
> drives only because that's what's on the back end.  If, however, you have a 
> 2530 and are connecting to it via a SAS HBA (which obviously you would be), 
> fcinfo will be of no help to you.
>
>   
>> We are trying to get MPxIO with SAS working with a
>> PCIe based SAS  
>> card.  I recently was told that if I have a PCIe
>> based SAS card with  
>>     
>>> 4gb of memory then I can't use MPxIO.  I'm hoping
>>>       
>>  that we get that  
>> ooner than later as it pretty much means I need to
>> (why is FC less  
>> than SAS for a 2530 anyways?) go with FC based
>> controllers.    Just  
>> wish that Sun could have been a little more upfront
>> about the "special  
>> situations" in which MPxIO+SAS wouldn't work.
>>     
>
> I believe what you're referring to is this CR:
>
> http://bugs.opensolaris.org/view_bug.do?bug_id=6597711
>
> The problem has to do with potential data corruption on DMA that crosses the 
> 4GB boundary.  This CR puts a workaround in the mpt driver to ensure that the 
> cards affected by this errata do not allocate DMA resources above 4GB.  If 
> you are running a driver with this CR, you're fine, although obviously there 
> could be performance degradation if you have a lot of physical memory.
>
> I'm sorry I don't have an answer for the original question of how to find the 
> LUN numbers.  I rarely have occasion to run with MPxIO enabled, and I'm not 
> too familiar with mpathadm.  Hopefully somebody else can answer that.
>
>   
A different LUN number can be assigned through LUN mapping for the same 
physical LU when mulitple target ports
exist. The mpathadm could show LUN number at path level but we had left 
up to transport layer utilities like fcinfo for reporting
it. There is no such utility for mpt but as far as LUN number is 
concerned, you can check 'prtconf -v' output and find lun prop from
pathinfo nodes of the scsi_vhci device.

--Hyon
> David
>  
>  
> This message posted from opensolaris.org
> _______________________________________________
> storage-discuss mailing list
> storage-discuss@opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/storage-discuss
>   

_______________________________________________
storage-discuss mailing list
storage-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to