[K12OSN] Linux "Software RAID"

[Les Mikesell]

Rob Owens wrote:
> 
>>
>> No, I don't believe so.  For one thing, as Dan Young put it, it's much 
>> easier to deal with swapping a failed disk out with a dedicated card.  
>> That by itself is a *BIG DEAL*.  Additionally, if you do have a disk 
>> fail, your CPU will take an especially big hit, because then it's got 
>> to reconstruct data from the parity info for *all* disk accesses, not 
>> just writes.  Oops....
>>
> I realize the ease-of-use advantage of hardware RAID, but I really was 
> talking about the performance hit only.

The performance hit mostly goes away if you use RAID1 instead of 5 (or 
the other direction - raid5 usually hurts performance unless you have 12 
or more drives in the set), but then you lose disk space.

> 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...

I'd think in terms of the difference in cost of software raid1 vs. 
hardware raid5 instead - that is, the extra disk space for full 
redundancy vs parity - and perhaps more drive bays in the server. If you 
have extra money in the mix, throwing in more RAM will almost always be 
a win in performance since it acts as a disk buffer and will eliminate 
the need for a lot of seeks and reads.

> Even though we've debated software vs. hardware RAID on this list many 
> times, these discussions always hold my interest.

The technology and pricing makes it a moving target.  A lot of 
server-class machines will have hardware RAID on the motherboard so it 
may not even be an extra-cost item, and some of the newer ones will mix 
and match SAS and SATA so you can balance performance and cost on the 
same hot-swap backplane.

-- 
   Les Mikesell
    lesmikesell at gmail.com