[OL] Re: Wanted - SATA/PATA/USB2 SSD with bad sectors for bcache testing (and offer of bcache presentation)

[Alan Johnson]

I don't think any of ours are bad, but I've got piles of 512GB SSDs at the
moment, and I'm about to order more, so if you would like to do some
benchmarking with lots of parallel SSDs, we might be able to help.  I don't
know if I can let them out of the building since they are supposed to go
into production eventually, but we have plenty of bench space and hardware
at the moment, so if you want to come here and play some day, just let me
know.

I am very interested in this project because we are currently paying Bill M.
some real money just to TRY and setup my "ultimate cloud storage solution"
use layers of Nexenta virtual machines to get ZFS introduced into our Ubuntu
friendly environment.  I am wondering if we could tackle this more simply
with bcache as part of the picture.  I also wonder how bcache would work
with FusionIO in the mix.  We are about to order a couple more of those, so
we should have a window where we can play with them before they go into
production.

Maybe you Bills and I should get together for lunch to see if there is some
way we can help eachother out.  Let me know.
_______________
Alan Johnson
alan at datdec.com


On Mon, May 10, 2010 at 1:46 PM, William Stearns <wstearns at pobox.com> wrote:

> Good afternoon, all,
>        "Seriously?  He wants _bad_ sectors?" I hear them say.  :-)
>
>        I'm working on a linux kernel project that uses SSD's to cache
> normal rotating media hard drives.  bcache is in early development and not
> stable for general use, but the performance numbers are enticing:
>
> * md5sum of small files: 3.5MB/sec off hard drive, 14.5MB/sec off SSD
> cache.
>
> * directory listing of large fragmented tree: 390 seconds off hard drive,
> 202 seconds off SSD cache.
>
> * md5sum of 1.995GB file with 23,522 extents: 251 seconds from HD
> (7.948MB/sec), 82 seconds from SSD cache (24.329MB/sec).
>
>        On a side note, I'd be happy to give a talk with a live demo on
> bcache to any gnhlug chapter; get in touch if you have an open month.
>
>        As part of the testing, I want to make sure that the code is
> resilient enough to recognize bad sectors on the SSD cache and not
> panic/bug/die.  That requires an SSD with bad sectors.  Size and speed are
> not critical.  I have the ability to connect to sata, pata, and usb.
>        Anyone have one?
>
>        Before I get inundated with offers, I already have a hard drive
> with underlying bad sectors for that part of the testing.  Come to think
> of it, I probably have 15 to 20 hard drives with.... issues.  :-)
>        Cheers,
>        - Bill
>
> ---------------------------------------------------------------------------
>        "Power concedes nothing without a demand. It never did, and it
> never will.  Find out just what people will submit to, and you have
> found out the exact amount of injustice and wrong which will be imposed
> upon them; and these will continue until they are resisted with either
> words or blows, or with both.  The limits of tyrants are prescribed by
> the endurance of those whom they oppress."
>        -- Frederick Douglass, August 4, 1857
> (Courtesy of Eric S. Raymond)
> --------------------------------------------------------------------------
> William Stearns (wstearns at pobox.com, tools and papers: www.stearns.org)
> Top-notch computer security training at www.sans.org , www.giac.net
> --------------------------------------------------------------------------
> _______________________________________________
> gnhlug-discuss mailing list
> gnhlug-discuss at mail.gnhlug.org
> http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.gnhlug.org/mailman/private/gnhlug-discuss/attachments/20100510/e66d9ae7/attachment.html