[linux-elitists] A modest proposal: SSD disk acceleration in kernel.

Karsten M. Self karsten@linuxmafia.com
Fri Oct 24 14:41:30 PDT 2008


on Fri, Oct 24, 2008 at 11:30:33AM -0400, Ruben Safir (ruben@mrbrklyn.com) wrote:
> On Fri, Oct 24, 2008 at 05:01:31PM +0200, Eugen Leitl wrote:
> > On Fri, Oct 24, 2008 at 10:58:40AM -0400, Ruben Safir wrote:
> > 
> > > BBad idea with current SSD's bcause of limited life.
> > 
> > Current SSDs have roughly enterprise drive reliability or better,
> > thanks to wear levelling algorithms. These are not your CFs.
> >  
> 
> 
> Eh - then it is time to ditch the platter drives for anything but the
> most intessive storage conditions.

As I noted:  there's still a near-order-of-magnitude cost differential.
For applications in which performance, durability, power, and/or weight
are a factor, SSDs will likely be (and already are) being substituted.

For vast, cheap, reliable (reasonably), non-performance-critical
applications, we'll be seeing rotating platters for quite some time.
I'm predicting at least a decade, possibly more.

Look at tape:  it's still not dead (though it's getting there).  It is
cheap, cheap, cheap (for media, not drives), and once you've got the
infrastructure in place, for structured backups, it's still a strong
contender.  The marginal cost of adding another generation of backups or
increment of storage is pretty low.  

Sure:  performance and convenience are a pain, but the cost wins out.

Good tech is hard to kill.  And with data needs exploding, I doubt SSD
will be sufficiently cheap to fit all needs for quite some time.


Peace.

-- 
Karsten M. Self <karsten@linuxmafia.com>        http://linuxmafia.com/~karsten
    Ceterum censeo, Caldera delenda est.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 191 bytes
Desc: Digital signature
Url : http://allium.zgp.org/pipermail/linux-elitists/attachments/20081024/c18cbabe/attachment.pgp 


More information about the linux-elitists mailing list