[Beowulf] SSD caching for parallel filesystems

Ellis H. Wilson III ellis at cse.psu.edu
Fri Feb 8 10:17:43 PST 2013

On 02/08/13 17:57, Mark Hahn wrote:
>> solid devices as well (also expensive however).  I think Micron also has
>> a native PCIe device in the wild now, the P320h?  Anybody know of other,
>> native PCIe devices?
> I'm not even sure what "native" PCIe flash would look like.  Do you mean
> that the driver and/or filesystem would have to do the wear-levelling and
> block remapping and garbage collection explicitly?

No, this is referring to the internal protocols of the SSD.  The SSD is 
just exposing a given protocol, but internally is managing many discrete 
storage devices (think baby RAID and micro-os in a box).

Right now, many SSD manufacturers are just SSD "repackagers" (including 
OCZ to my knowledge).  They buy a controller design from one place (some 
make this component), SSD packages from someplace else, some channel 
controllers, etc, etc, and strap it all together.  Which is totally 
fine, but the problem arises because the volume for NAND flash packages 
are for SATA based drives.  This results in most of the NAND packages 
within to export a SATA protocol.  This requires these re-packaging 
companies to have to translate to and from the SATA and PCIe protocols. 
  For another explanation, please see the fourth paragraph of:


Hopefully this explains better the issue I'm referring to.

Managing the raw flash at the filesystem or driver level is an option, 
but not what I was talking about (nor what is currently in vogue right 
now, to my knowledge).

To be fair, I do like OCZ's solutions for retail SATA drives.  But the 
jump from that protocol to PCIe, if you are going to make it, needs to 
be made with some consideration.  It's more expensive and is a more 
diverse space in terms of "real" performance provided by the various 
vendors.  In even a year or two this probably won't be the case, but 
right now there is a lot of junk on the market.

Just my 2c though,


More information about the Beowulf mailing list