Re: [PLUG] swappiness and ssd

I was actually tangentially connected to that testing, and while it generally had large non-compressable chunks written across the whole disk, there were interval tests that involved writing particular groups of smaller files to the SSDs and reading them back. Those actually showed issues a bit before the big overwrite tests, so Rich Prime noted, I wouldn't extrapolate those tests to mean that you can take any given consumer drive and beat it up with a few hundred TB of writes and then be shocked when it dies.

I've been watching use on my Samsung 830 Pro 256GB at home for quite a few years now, and there are definitely 'hot spots' of intense activity. I'll be planning to replace my SSD at around the 250TB mark, and I set swappiness to 0.

All FWIW.

On Thu, Apr 16, 2015 at 12:54 PM, Soren Harward <stharward@gmail.com> wrote:

On Thu, Apr 16, 2015 at 10:24 AM, Anthony Martin
<anthony.j.martin142@gmail.com> wrote:
> Also as both Rich and Rich said the small performance increase you
> would get isn't really worth the damage to you SSD.

Recent testing stats [0] 've seen are that consumer-grade SSDs are
good for at least 500TB of writes, and many survive well past 2PB. So
they're unlikely to be significantly degraded even by using them for a
swap drive. Unless you've got so little swap you're continuously
paging in and out, which case you should spend the money on RAM
instead.

[0] http://techreport.com/review/27436/the-ssd-endurance-experiment-two-freaking-petabytes

--
Soren Harward

___________________________________________________________________________
Philadelphia Linux Users Group -- http://www.phillylinux.org
Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce
General Discussion -- http://lists.phillylinux.org/mailman/listinfo/plug