Date: Sun, 24 Dec 2023 11:50:07 +0000 From: void <void@f-m.fm> To: freebsd-fs@freebsd.org Subject: Re: measuring swap partition speed Message-ID: <ZYga78SOTI4LrimJ@int21h> In-Reply-To: <B4833CC9-96FB-444E-B1BC-2662A2909E7A@yahoo.com> References: <B4833CC9-96FB-444E-B1BC-2662A2909E7A.ref@yahoo.com> <B4833CC9-96FB-444E-B1BC-2662A2909E7A@yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Dec 23, 2023 at 11:55:43AM -0800, Mark Millard wrote:
>That last indicates another potentially significant difference
>from my example benchmarks done for comparison to yours:
>
>GEOM_ELI: Device da0p3.eli created.
>GEOM_ELI: Encryption: AES-XTS 256
>GEOM_ELI: Crypto: software
>
>That likely contributes to slowing down I/O for da0p3 and
>possibly to leaving less time for I/O to other partitions
>when da0p3 is in use.
I think *some* slowdown is reasonable. What's happening here, though,
amounts to blocking. 300-1000Kb/s against 16MB.
zfs filesystem reads/writes are slower to an extent but it's
not by a massive amount, like it is with swap, (and the swap
is unencrypted)
>If you had the resources to test avoiding the encrypted
>partition for your type of media, that might prove
>interesting.
I think i did early on but I can do so again. I'll delete
some swap partitions and format as ufs.
>I've never had occasion to want to use any encrypted
>partitions for my FreeBSD activity.
A couple of reasons for this particular instance:
1. the hardware is portable and visible and would be easily
stolen by someone of a mind to
2. the system performs authentication of which some parts
are in clear txt
>I'm unsure of the GEOM_ELI processing-time contributions
>vs. "gstat -spod" reporting.
yeah I've long wanted to measure the performance penalty for this,
given that theres no aesni or equivalent capability available in the
hardware.
--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?ZYga78SOTI4LrimJ>