[K12OSN] Linux "Software RAID"

[Dan Young]

On Fri, Aug 8, 2008 at 10:16 AM, Rob Owens
<rob.owens at biochemfluidics.com> wrote:
> Dan Young wrote:
>>
>> On Fri, Aug 8, 2008 at 4:37 AM, Rob Owens <rob.owens at biochemfluidics.com>
>> wrote:
>>>
>>> I realize the ease-of-use advantage of hardware RAID, but I really was
>>> talking about the performance hit only.
>>>
>>> Realistically, a hardware RAID costs about $300, and that could buy me a
>>> 2nd
>>> 4-core processor.  I have to believe that a modern 4-core processor is
>>> way
>>> more capable of handling a RAID rebuild than whatever chip is onboard a
>>> hardware RAID controller.  Of course, I haven't done any testing...
>>
>> If you have a md on Linux running somewhere, try "dmesg | grep raid".
>> You'll see that the kernel tests several checksum algorithms, some
>> using SIMD ISA extensions:
>>
>> raid5: automatically using best checksumming function: pIII_sse
>>   pIII_sse  :  2604.000 MB/sec
>> raid5: using function: pIII_sse (2604.000 MB/sec)
>> raid6: int32x1    312 MB/s
>> raid6: int32x2    375 MB/s
>> raid6: int32x4    351 MB/s
>> raid6: int32x8    339 MB/s
>> raid6: mmxx1     1023 MB/s
>> raid6: mmxx2     1238 MB/s
>> raid6: sse1x1     921 MB/s
>> raid6: sse1x2    1261 MB/s
>> raid6: using algorithm sse1x2 (1261 MB/s)
>>
> I don't see any of that info on my systems with software RAID 1.

There's no parity calculation for RAID 1, it's just a mirror copy.

> What do
> those values mean?  Are they read speads?  Write speeds?

I would presume it's the speed at which the processor can XOR your data.

--
Dan Young <dyoung at mesd.k12.or.us>
Multnomah ESD - Technology Services
503-257-1562