freebsd-fs Digest, Vol 630, Issue 5

[Pokala, Ravi]

>Date: Sat, 8 Aug 2015 14:41:07 +0300
>From: Konstantin Belousov <kostikbel at gmail.com>
>To: Willem Jan Withagen <wjw at digiware.nl>
>Cc: fs at freebsd.org
>Subject: Re: Using SSDs as swap
>Message-ID: <20150808114107.GD2072 at kib.kiev.ua>
>Content-Type: text/plain; charset=us-ascii
>
>> 
>> Speaking about blocks.... Does Swap take into account that disks could
>>be of a sectorsize other than 512 bytes. I would guess so, since we
>>could have a 4K disk as swap disk, and doing read-modify-write for swap
>>is sure going to kill performance.
>swap performs i/o in the page-sized chunks at least, which are min 4k on
>all supported platforms (even on arms, where we do not support smaller
>pages AFAIK).

I can confirm this first-hand - I have systems which use AF-4Kn drives for
swap, without any problems.

HOWEVER, there is a different - but related - snag. Frequently, the same
device is used for both swap and vmcores. While swapping (aka paging) is
done in 4KB chunks, some structures used in dumping core are always 512B.
This does not work well on AF-4Kn drives. [PR 194279] discusses the
problem.

cem@ did some work on the dump path "recently" (yikes - according to the
bug, it was close to a year ago!)

-Ravi