Dev, Vasu wrote:
>> -----Original Message-----
>> From: Mike Christie [mailto:[EMAIL PROTECTED]
>> Yeah, I am just saying if we have to add complexity to handle it it
>> might not be worth it because at this point performance is hosed. If we
>> were failing the IO with different errors codes to try and get it
> failed
>> over for multipath quickly that might be different. But if we can
> simply
>> check the reject response and reason and check what is going on with
> the
>> command, it should be fine (not complex) and I agree with you that we
>> should do it for certain safe cases.
>>
> 
> I don't know those all certain safe cases but yes it will be a simple
> change to check very few  BA_RJT reason and explanations codes here for
> safe cases. And I also don't know most likely those safe cases and their
> typical frequency to assess worth of handling some of the BA_RJT codes
> here v/s let always scsi-eh recover from such cases. 
> 
> However I think BA_RJT with FC_BA_RJT_NONE and FC_BA_RJT_LOG_ERR(with
> FC_BA_RJT_INV_XID) will be most common safe case in poor quality lossy
> link to complete exch here in FCP to avoid lengthy scsi-eh in this case
> to all the way to lun reset as you described earlier in this email as
> per current code state. I can complete exch here for only these cases
> (FC_BA_RJT_NONE & FC_BA_RJT_LOG_ERR) since at least I see BA_RJT with
> FC_BA_RJT_LOG_ERR and FC_BA_RJT_INV_XID quite often when a scsi response
> is dropped to my target and FC_BA_RJT_NONE seem general case to assume
> safe case here. I want to avoid lun reset for at least these cases as
> per current code.
> 

Sounds good to me.
_______________________________________________
devel mailing list
[email protected]
http://www.open-fcoe.org/mailman/listinfo/devel

Reply via email to