Rich Freeman on 12 Jul 2015 12:49:57 -0700

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: [PLUG] swappiness and ssd

On Sun, Jul 12, 2015 at 10:14 AM, K.S. Bhaskar <> wrote:
> The traditional view of swapping is that it is undesirable because reading
> from swap is slow, and swap should therefore be used only if unavoidable.
> Therefore, the received wisdom is to set a low value of swappiness, to swap
> only when unavoidable, and take a performance hit when that happened.
> But, write cycle limits aside, I believe SSDs require a rethink.

I'd be very interested in feedback from those using swap on SSD.
However, I never thought that swapiness=0 was conventional wisdom on
Linux.  I can think of a million reasons why a higher setting /should/
be better.  The problem is that it often isn't, and that is just due
to limitations in Linux and how it is used.  The defaults probably
make running updatedb twice in a row a lot faster, but the problem is
that nobody actually does that, and all that swapping after running it
once kills everything else you do.

It has been a while since I've run with swap, so it is possible that
things have gotten better.  I do agree that performance of swap should
be much better on an SSD.

The other thing I'd be concerned with is ssd write cycles.  I've tried
to move a lot of my heavy-modification activities off of ssd for this
reason.  Swap is going to tend to wear it faster.

Philadelphia Linux Users Group         --
Announcements -
General Discussion  --