From: Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>
To: Reindl Harald <h.reindl@thelounge.net>
Cc: Piergiorgio Sartor <piergiorgio.sartor@nexgo.de>,
linux-raid@vger.kernel.org
Subject: Re: RAID-10 near vs. RAID-1
Date: Wed, 12 Jun 2024 19:22:42 +0200 [thread overview]
Message-ID: <ZmnZYgerX5g8S9Cp@lazy.lzy> (raw)
In-Reply-To: <cd3ed227-1410-478b-b86b-973d76b587df@thelounge.net>
On Wed, Jun 12, 2024 at 01:04:18AM +0200, Reindl Harald wrote:
>
>
> Am 11.06.24 um 20:31 schrieb Piergiorgio Sartor:
> > I'm setting up a system with 2 SSD M.2 (NVME).
> >
> > I was wondering if would it be better, performace
> > wise, to have a RAID-10 near layout or a RAID-1.
> >
> > Looking around I found only one benchmark:
> >
> > https://strugglers.net/~andy/blog/2019/06/02/exploring-different-linux-raid-10-layouts-with-unbalanced-devices/
> >
> > Which uses mixed SSD, NVME and SATA.
> >
> > Does anybody have any suggestions, links, or
> > ideas on the topic?
> >
> > BTW, practically speaking, what's the difference,
> > between the two RAIDs?
>
> i wouldn't even consider a RAID10 with two disks, especially with SSD and
> practically you end with a unsupported RAID1 because there are no stripes
> with 2 disks
>
Hi, thanks for the answer.
I'm a bit confused here. What do you mean
with "unsupported RAID1"?
As far as I know, but please correct me if
I'm wrong, a Linux md RAID-10 *near* layout,
with 2 devices, has identical data distribution
as a RAID-1 with 2 devices.
Meaning the 2 devices are a mirror.
The difference, if I understood it correctly,
is that the RAID-10 has chunks, and hence stripes,
while the RAID-1 does not have stripes.
Furthermore, the read operation on RAID-10 are
interleaved, delivering (for SSDs) double
sequential read speed (for 2 devices), while
the RAID-1 can handle two independent (one per
device) read stream, each with single device
reading speed.
Of course, depending on the requirements,
assuming what I wrote is correct, performances
might be different.
I was just wondering if anybody has some hints,
some experience, some references, or, as you
suggested, not to care at all.
Thanks again,
bye,
--
piergiorgio
next prev parent reply other threads:[~2024-06-12 17:30 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-11 18:31 RAID-10 near vs. RAID-1 Piergiorgio Sartor
2024-06-11 23:04 ` Reindl Harald
2024-06-12 0:14 ` Paul E Luse
2024-06-12 7:06 ` Reindl Harald
2024-06-12 7:18 ` Reindl Harald
2024-06-12 17:25 ` Piergiorgio Sartor
2024-06-12 17:22 ` Piergiorgio Sartor [this message]
2024-06-12 22:35 ` Reindl Harald
2024-06-12 23:46 ` Dragan Milivojević
2024-06-13 5:38 ` Reindl Harald
2024-06-13 7:30 ` Robin Hill
[not found] ` <CALtW_agtMXsss_Y=A2HH+D5zTceJ0jv5eWM5OeKiRZphvVeXZw@mail.gmail.com>
[not found] ` <599595a2-fa5e-45ca-b358-5fb573a8920e@thelounge.net>
2024-06-13 19:54 ` Dragan Milivojević
2024-06-13 20:18 ` Reindl Harald
[not found] ` <CALtW_ageds8cA-3CgbSNW5sFmRvWGmqoM0vA1vbi5LxWLhgt7g@mail.gmail.com>
2024-06-13 20:53 ` Reindl Harald
2024-06-25 19:00 ` Phillip Susi
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZmnZYgerX5g8S9Cp@lazy.lzy \
--to=piergiorgio.sartor@nexgo.de \
--cc=h.reindl@thelounge.net \
--cc=linux-raid@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox